Release processed Focus mat to not leak, cache, and fix cvmat refcounting (#2356)

## Description

Presently, any tests using our old configs happen in place. This is
problematic, as it changes the files themselves. This means that anyone
running the tests will cause unintentional modifications, which stand a
chance of being committed and merged into main. We want these old
database files to remain untouched, thus we copy them to a temp
directory prior to running our tests.

## Meta

Merge checklist:
- [x] Pull Request title is [short, imperative
summary](https://cbea.ms/git-commit/) of proposed changes
- [x] The description documents the _what_ and _why_, including events
that led to this PR
- [ ] If this PR changes behavior or adds a feature, user documentation
is updated
- [ ] If this PR touches photon-serde, all messages have been
regenerated and hashes have not changed unexpectedly
- [ ] If this PR touches configuration, this is backwards compatible
with all settings going back to the previous seasons's last release
(seasons end after champs ends)
- [ ] If this PR touches pipeline settings or anything related to data
exchange, the frontend typing is updated
- [ ] If this PR addresses a bug, a regression test for it is added

.github/pull_request_template.md

@@ -1,18 +1,18 @@
 ## Description
 
-<!-- What changed? Why? (the code + comments should speak for itself on the "how") -->
+What changed? Why? (the code + comments should speak for itself on the "how")
 
-<!-- Fun screenshots or a cool video or something are super helpful as well. If this touches platform-specific behavior, this is where test evidence should be collected. -->
+Include fun testing screenshots or a cool video, to collect test evidence in a place where we can later reference it. Including proof this change was tested makes reviewing easier, helps us make sure we tested all our edge cases, and helps provide context for the future.
 
-<!-- Any issues this pull request closes or pull requests this supersedes should be linked with `Closes #issuenumber`. -->
+Any issues this pull request closes or pull requests this supersedes should be linked with `Closes #issuenumber`.
 
 ## Meta
 
 Merge checklist:
 - [ ] Pull Request title is [short, imperative summary](https://cbea.ms/git-commit/) of proposed changes
-- [ ] The description documents the _what_ and _why_
+- [ ] The description documents the _what_ and _why_, including events that led to this PR
 - [ ] If this PR changes behavior or adds a feature, user documentation is updated
 - [ ] If this PR touches photon-serde, all messages have been regenerated and hashes have not changed unexpectedly
-- [ ] If this PR touches configuration, this is backwards compatible with settings back to v2025.3.2
+- [ ] If this PR touches configuration, this is backwards compatible with all settings going back to the previous seasons's last release (seasons end after champs ends)
 - [ ] If this PR touches pipeline settings or anything related to data exchange, the frontend typing is updated
 - [ ] If this PR addresses a bug, a regression test for it is added

.github/workflows/build.yml

@@ -10,7 +10,7 @@ concurrency:
   cancel-in-progress: true
 
 env:
-  IMAGE_VERSION: v2026.1.1
+  IMAGE_VERSION: v2026.1.2
 
 jobs:
 
@@ -555,32 +555,32 @@ jobs:
           minimum_free_mb: ${{ matrix.minimum_free_mb }}
           root_location: ${{ matrix.root_location || 'partition=2' }}
           shrink_image: ${{ matrix.shrink_image || 'yes' }}
-          commands: |
-            chmod +x scripts/armrunner.sh
-            ./scripts/armrunner.sh
-            java -jar *.jar --smoketest --platform=${{ matrix.plat_override }}
+          commands: ./scripts/armrunner.sh
 
       - name: Compress image
         # Compress the standard images
         if: ${{ ! startsWith(matrix.image_suffix, 'rubik') }}
         run: |
+          set -ex
           new_jar=$(realpath $(find . -name photonvision\*-linuxarm64.jar))
           new_image_name=$(basename "${new_jar/.jar/_${{ matrix.image_suffix }}.img}")
           sudo mv ${{ steps.generate_image.outputs.image }} $new_image_name
-          sudo xz -T 0 -v $new_image_name
+          sudo xz -T 0 -kv $new_image_name
+          echo "smoketest_image_loc=${new_image_name}" >> $GITHUB_ENV
 
-      - name: Tar built image
+      - name: Tar built image (Rubik)
         # Build the RubikPi3-specific tar file
         if: ${{ startsWith(matrix.image_suffix, 'rubik') }}
         run: |
+          set -ex
           new_jar=$(realpath $(find . -name photonvision\*-linuxarm64.jar))
-          new_image_name=$(basename "${new_jar/.jar/_${{ matrix.image_suffix }}.img}")
-
+          tardir=$(basename "${new_jar/.jar/_${{ matrix.image_suffix }}.img}")
           imagedir=$(dirname ${{ steps.generate_image.outputs.image }})
-          tardir=${new_image_name}
           sudo mkdir --parents ${tardir}
-          sudo mv ${imagedir}/* ${tardir}/
-          sudo tar -I 'xz -T0' -cf ${new_image_name}.tar.xz ${tardir} --checkpoint=10000 --checkpoint-action=echo='%T'
+          sudo cp ${imagedir}/* ${tardir}/
+          sudo tar -I 'xz -T0' -cf ${tardir}.tar.xz ${tardir} --checkpoint=10000 --checkpoint-action=echo='%T'
+          # Point smoketest to the old image
+          echo "smoketest_image_loc=${{ steps.generate_image.outputs.image }}" >> $GITHUB_ENV
 
       - uses: actions/upload-artifact@v6
         name: Upload image
@@ -588,6 +588,16 @@ jobs:
           name: image-${{ matrix.image_suffix }}
           path: photonvision*.xz
 
+        # This is done after uploading the image to avoid contaminating the image with logs, caches, etc.
+      - uses: photonvision/photon-image-runner@HEAD
+        name: Smoketest Image
+        with:
+          image_url: file://${{ env.smoketest_image_loc }}
+          minimum_free_mb: ${{ matrix.minimum_free_mb }}
+          root_location: ${{ matrix.root_location || 'partition=2' }}
+          shrink_image: ${{ matrix.shrink_image || 'yes' }}
+          commands: java -jar *.jar --smoketest --platform=${{ matrix.plat_override }}
+
   matrix-checker:
     # This job always runs last to set the overall result based on the matrix jobs. If any matrix job failed, this job will fail.
     # This makes it so that we don't need to add each matrix job individually to CI checks.

.github/workflows/python.yml

@@ -12,7 +12,7 @@ concurrency:
   cancel-in-progress: true
 
 jobs:
-  build-and-deploy:
+  build-py:
     runs-on: ubuntu-24.04
 
     steps:
@@ -29,38 +29,57 @@ jobs:
       - name: Install dependencies
         run: |
           python -m pip install --upgrade pip
-          pip install setuptools wheel pytest mypy
+          pip install setuptools wheel
 
       - name: Build wheel
         working-directory: ./photon-lib/py
         run: python setup.py sdist bdist_wheel
 
-      - name: Run Unit Tests
-        working-directory: ./photon-lib/py
-        run: |
-            pip install --no-cache-dir dist/*.whl
-            pytest
-
-      - name: Run mypy type checking
-        run: mypy --show-column-numbers --config-file photon-lib/py/pyproject.toml photon-lib
-
       - name: Upload artifacts
         uses: actions/upload-artifact@v6
         with:
           name: dist
           path: ./photon-lib/py/dist/
 
-      - name: Publish package distributions to TestPyPI
-        # Only upload on tags
-        if: startsWith(github.ref, 'refs/tags/v')
-        uses: pypa/gh-action-pypi-publish@release/v1
-        with:
-          packages-dir: ./photon-lib/py/dist/
+  test-py:
+    needs: build-py
+    runs-on: ubuntu-24.04
 
-    permissions:
-      id-token: write  # IMPORTANT: this permission is mandatory for trusted publishing
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v6
+        with:
+          fetch-depth: 0
+
+      - name: Set up Python
+        uses: actions/setup-python@v6
+        with:
+          python-version: 3.14
+
+      - name: Install dependencies
+        run: |
+          python -m pip install --upgrade pip
+          pip install pytest mypy
+
+      - name: Download artifacts
+        uses: actions/download-artifact@v6
+        with:
+          name: dist
+          path: dist/
+
+      - name: Install package
+        shell: bash
+        run: pip install --no-cache-dir dist/*.whl
+
+      - name: Run Unit Tests
+        shell: bash
+        run: pytest --import-mode=importlib photon-lib/py/test/
+
+      - name: Run mypy type checking
+        run: mypy --show-column-numbers --config-file photon-lib/py/pyproject.toml photon-lib
 
   build-python-examples:
+    needs: build-py
     strategy:
       matrix:
         os: [ubuntu-24.04, windows-2022, macos-14]
@@ -81,28 +100,18 @@ jobs:
       - name: Install dependencies
         run: |
           python -m pip install --upgrade pip
-          pip install setuptools wheel pytest mypy
 
-      - name: Build wheel
-        working-directory: ./photon-lib/py
-        run: python setup.py sdist bdist_wheel
+      - name: Download artifacts
+        uses: actions/download-artifact@v6
+        with:
+          name: dist
+          path: ./photon-lib/py/dist/
 
-      - name: Build and configure PhotonLibPy
-        working-directory: ./photon-lib/py
-        shell: bash
-        run: |
-          ./buildAndTest.sh
-          ./enableUsingDevBuilds.sh
-
-      - name: Run Unit Tests
+      - name: Install PhotonLibPy package
         working-directory: ./photon-lib/py
         shell: bash
         run: |
             pip install --no-cache-dir dist/*.whl
-            pytest
-
-      - name: Run mypy type checking
-        run: mypy --show-column-numbers --config-file photon-lib/py/pyproject.toml photon-lib
 
       - name: Build Python examples
         working-directory: photonlib-python-examples
@@ -113,3 +122,24 @@ jobs:
             echo $folder
             ./run.sh $folder
           done
+
+  deploy:
+    needs: [test-py, build-python-examples]
+    runs-on: ubuntu-24.04
+    # Only upload on tags
+    if: startsWith(github.ref, 'refs/tags/v')
+
+    steps:
+      - name: Download artifacts
+        uses: actions/download-artifact@v6
+        with:
+          name: dist
+          path: dist/
+
+      - name: Publish package distributions to PyPI
+        uses: pypa/gh-action-pypi-publish@release/v1
+        with:
+          packages-dir: ./dist/
+
+    permissions:
+      id-token: write  # IMPORTANT: this permission is mandatory for trusted publishing

README.md

@@ -8,18 +8,18 @@ The latest release of platform-specific jars and images is found [here](https://
 
 If you are interested in contributing code or documentation to the project, please [read our getting started page for contributors](https://docs.photonvision.org/en/latest/docs/contributing/index.html) and **[join the Discord](https://discord.gg/wYxTwym) to introduce yourself!** We hope to provide a welcoming community to anyone who is interested in helping.
 
-## Authors
-
-<a href="https://github.com/PhotonVision/photonvision/graphs/contributors">
-  <img src="https://contrib.rocks/image?repo=PhotonVision/photonvision" />
-</a>
-
 ## Documentation
 
 - Our main documentation page: [docs.photonvision.org](https://docs.photonvision.org)
 - Photon UI demo: [demo.photonvision.org](https://demo.photonvision.org)
 - Javadocs: [javadocs.photonvision.org](https://javadocs.photonvision.org)
-- C++ Doxygen  [cppdocs.photonvision.org](https://cppdocs.photonvision.org)
+- C++ Doxygen: [cppdocs.photonvision.org](https://cppdocs.photonvision.org)
+
+## Authors
+
+<a href="https://github.com/PhotonVision/photonvision/graphs/contributors">
+  <img src="https://contrib.rocks/image?repo=PhotonVision/photonvision" />
+</a>
 
 ## Building
 
@@ -32,7 +32,6 @@ You can run one of the many built in examples straight from the command line, to
 Note that these are case sensitive!
 
 * `-PArchOverride=foobar`: builds for a target system other than your current architecture. [Valid overrides](https://github.com/wpilibsuite/wpilib-tool-plugin/blob/main/src/main/java/edu/wpi/first/tools/NativePlatforms.java) are:
-    * winx32
     * winx64
     * winarm64
     * macx64
@@ -46,33 +45,34 @@ Note that these are case sensitive!
 - `-Pprofile`: enables JVM profiling
 - `-PwithSanitizers`: On Linux, enables `-fsanitize=address,undefined,leak`
 
-If you're cross-compiling, you'll need the wpilib toolchain installed. This can be done via Gradle: for example `./gradlew installArm64Toolchain` or `./gradlew installRoboRioToolchain`
+If you're cross-compiling, you'll need the WPILib toolchain installed. This must be done via Gradle: for example `./gradlew installArm64Toolchain` or `./gradlew installRoboRioToolchain`
 
 ## Out-of-Source Dependencies
 
 PhotonVision uses the following additional out-of-source repositories for building code.
 
-- Base system images for Raspberry Pi & Orange Pi: https://github.com/PhotonVision/photon-image-modifier
+- Base system images for supported coprocessors: https://github.com/PhotonVision/photon-image-modifier
 - C++ driver for Raspberry Pi CSI cameras: https://github.com/PhotonVision/photon-libcamera-gl-driver
 - JNI code for [mrcal](https://mrcal.secretsauce.net/): https://github.com/PhotonVision/mrcal-java
-- Custom build of OpenCV with GStreamer/Protobuf/other custom flags: https://github.com/PhotonVision/thirdparty-opencv
-- JNI code for aruco-nano: https://github.com/PhotonVision/aruconano-jni
+- JNI code for RKNN: https://github.com/PhotonVision/rknn_jni
+- JNI code for Rubik Pi NPU: https://github.com/PhotonVision/rubik_jni
 
 ## Acknowledgments
 
 PhotonVision was forked from [Chameleon Vision](https://github.com/Chameleon-Vision/chameleon-vision/). Thank you to everyone who worked on the original project.
 
-* [WPILib](https://github.com/wpilibsuite) - Specifically [cscore](https://github.com/wpilibsuite/allwpilib/tree/main/cscore), [CameraServer](https://github.com/wpilibsuite/allwpilib/tree/main/cameraserver), [NTCore](https://github.com/wpilibsuite/allwpilib/tree/main/ntcore), and [OpenCV](https://github.com/wpilibsuite/thirdparty-opencv).
-
-* [Apache Commons](https://commons.apache.org/) - Specifically [Commons Math](https://commons.apache.org/proper/commons-math/), and [Commons Lang](https://commons.apache.org/proper/commons-lang/)
-
+* [WPILib](https://github.com/wpilibsuite) - Specifically [allwpilib](https://github.com/wpilibsuite/allwpilib) and [their build of OpenCV](https://github.com/wpilibsuite/thirdparty-opencv).
+* [Apache Commons](https://commons.apache.org/) - Specifically [Commons IO](https://commons.apache.org/proper/commons-io/), and [Commons CLI](https://commons.apache.org/proper/commons-cli/)
+* [diozero](https://www.diozero.com/)
+* [EJML](https://github.com/lessthanoptimal/ejml)
 * [Javalin](https://javalin.io/)
-
 * [JSON](https://json.org)
-
 * [FasterXML](https://github.com/FasterXML) - Specifically [jackson](https://github.com/FasterXML/jackson)
-
+* [MessagePack for Java](https://github.com/msgpack/msgpack-java)
 * [OSHI](https://github.com/oshi/oshi)
+* [QuickBuffers](https://github.com/HebiRobotics/QuickBuffers)
+* [SQLite JDBC](https://github.com/xerial/sqlite-jdbc)
+* [ZT ZIP](https://github.com/zeroturnaround/zt-zip)
 
 ## License
 

build.gradle

@@ -4,7 +4,7 @@ plugins {
     id "cpp"
     id "com.diffplug.spotless" version "8.1.0"
     id "edu.wpi.first.wpilib.repositories.WPILibRepositoriesPlugin" version "2020.2"
-    id "edu.wpi.first.GradleRIO" version "2026.1.1"
+    id "edu.wpi.first.GradleRIO" version "2026.2.1"
     id 'org.photonvision.tools.WpilibTools' version '2.3.3-photon'
     id 'com.google.protobuf' version '0.9.3' apply false
     id 'edu.wpi.first.GradleJni' version '1.1.0'
@@ -32,7 +32,7 @@ ext.allOutputsFolder = file("$project.buildDir/outputs")
 apply from: "versioningHelper.gradle"
 
 ext {
-    wpilibVersion = "2026.1.1"
+    wpilibVersion = "2026.2.1"
     wpimathVersion = wpilibVersion
     openCVYear = "2025"
     openCVversion = "4.10.0-3"

docs/source/docs/advanced-installation/sw_install/files/Limelight2+/hardwareConfig.json

@@ -3,6 +3,6 @@
   "supportURL" : "https://limelightvision.io",
   "ledPins" : [ 13, 18 ],
   "ledsCanDim" : true,
-  "ledPWMFrequency" : 30000,
+  "ledPWMFrequency" : 1000,
   "vendorFOV" : 75.76079874010732
 }

docs/source/docs/calibration/calibration.md

@@ -83,7 +83,7 @@ Details about a particular calibration can be viewed by clicking on that resolut
 More info on what these parameters mean can be found in [OpenCV's docs](https://docs.opencv.org/4.8.0/d4/d94/tutorial_camera_calibration.html)
 :::
 
-- Fx/Fy: Estimated camera focal length, in mm
+- Fx/Fy: Estimated camera focal length, in pixels
 - Fx/Cy: Estimated camera optical center, in pixels. This should be at about the center of the image
 - Distortion: OpenCV camera model distortion coefficients
 - FOV: calculated using estimated focal length and image size. Useful for gut-checking calibration results

docs/source/docs/examples/poseest.md

@@ -18,7 +18,7 @@ The {code}`Drivetrain` class includes functionality to fuse multiple sensor read
 
 Please reference the [WPILib documentation](https://docs.wpilib.org/en/stable/docs/software/advanced-controls/state-space/state-space-pose_state-estimators.html) on using the {code}`SwerveDrivePoseEstimator` class.
 
-We use the 2024 game's AprilTag Locations:
+We use the current game's AprilTag Locations:
 
 ```{eval-rst}
 .. tab-set-code::

docs/source/docs/hardware/customhardware.md

@@ -73,27 +73,21 @@ If you were using custom LED commands from 2025 or earlier and still need custom
 
 ## Hardware Interaction Commands
 
-For Non-Raspberry-Pi hardware, users must provide valid hardware-specific commands for some parts of the UI interaction (including performance metrics, and executing system restarts).
+For non-Linux hardware, users must provide the hardware-specific command for executing system restarts.
 
-Leaving a command blank will disable the associated functionality.
+Leaving this command blank will disable the restart functionality.
 
 ```{eval-rst}
 .. tab-set-code::
    .. code-block::  json
 
       {
-        "cpuTempCommand" : "",
-        "cpuMemoryCommand" : "",
-        "cpuUtilCommand" : "",
-        "gpuMemoryCommand" : "",
-        "gpuTempCommand" : "",
-        "ramUtilCommand" : "",
         "restartHardwareCommand" : "",
       }
 ```
 
 :::{note}
-These settings have no effect if PhotonVision detects it is running on a Raspberry Pi. See [the MetricsBase class](https://github.com/PhotonVision/photonvision/blob/dbd631da61b7c86b70fa6574c2565ad57d80a91a/photon-core/src/main/java/org/photonvision/common/hardware/metrics/MetricsBase.java) for the commands utilized.
+This setting has no effect if PhotonVision detects it is running on Linux. On Linux, the restart is accomplished by executing `reboot now` in a shell.
 :::
 
 ## Known Camera FOV
@@ -150,13 +144,7 @@ Here is a complete example `hardwareConfig.json`:
         "setGPIOCommand" : "setGPIO {p} {s}",
         "setPWMCommand" : "setPWM {p} {v}",
         "setPWMFrequencyCommand" : "setPWMFrequency {p} {f}",
-        "releaseGPIOCommand" : "releseGPIO {p}",
-        "cpuTempCommand" : "",
-        "cpuMemoryCommand" : "",
-        "cpuUtilCommand" : "",
-        "gpuMemoryCommand" : "",
-        "gpuTempCommand" : "",
-        "ramUtilCommand" : "",
+        "releaseGPIOCommand" : "releaseGPIO {p}",
         "restartHardwareCommand" : "",
         "vendorFOV" : 72.5
       }

docs/source/docs/objectDetection/about-object-detection.md

@@ -4,7 +4,7 @@
 
 PhotonVision supports object detection using neural network accelerator hardware, commonly known as an NPU. The two coprocessors currently supported are the {ref}`Orange Pi 5 <docs/objectDetection/opi:Orange Pi 5 (and variants) Object Detection>` and the {ref}`Rubik Pi 3 <docs/objectDetection/rubik:Rubik Pi 3 Object Detection>`.
 
-PhotonVision currently ships with a model trained on the [COCO dataset](https://cocodataset.org/) by [Ultralytics](https://github.com/ultralytics/ultralytics) (this model is licensed under [AGPLv3](https://www.gnu.org/licenses/agpl-3.0.en.html)). This model is meant to be used for testing and other miscellaneous purposes. It is not meant to be used in competition. For the 2025 post-season, PhotonVision also ships with a pretrained ALGAE model. A model to detect coral is available in the PhotonVision discord, but will not be distributed with PhotonVision.
+PhotonVision currently ships with a model trained on the [COCO dataset](https://cocodataset.org/) by [Ultralytics](https://github.com/ultralytics/ultralytics) (this model is licensed under [AGPLv3](https://www.gnu.org/licenses/agpl-3.0.en.html)). This model is meant to be used for testing and other miscellaneous purposes. It is not meant to be used in competition. For the 2026 season, PhotonVision ships with a model to detect FUEL, this is also licensed under AGPL.
 
 ## Tracking Objects
 

docs/source/docs/programming/photonlib/robot-pose-estimator.md

@@ -108,7 +108,7 @@ When taking in a result from a `PhotonCamera`, PhotonPoseEstimator offers nine p
     flat on the floor. This computation takes place on the RoboRIO, and should not take more than 2ms.
     This also requires addHeadingData to be called every frame so heading data is up to date.
 
-Calling one of the `estimate<strategy>Pose()` methods on your `PhotonPoseEstimator` will return an `Optional<EstimatedRobotPose>`, which includes a `Pose3d` of the latest estimated pose (using the selected strategy) along with a `double` of the timestamp when the robot pose was estimated. The recommended way to use the estimatePose methods is to
+Calling one of the `estimate<strategy>Pose()` methods on your `PhotonPoseEstimator` will return an `Optional<EstimatedRobotPose>`, which will be empty if there are no detected tags, not enough detected tags (for multi-tag strategies), missing data (typically heading data), or if the internal solvers failed (this is a rare scenario). `EstimatedRobotPose` includes a `Pose3d` of the latest estimated pose (using the selected strategy) along with a `double` of the timestamp when the robot pose was estimated. The recommended way to use the estimatePose methods is to
 1. do estimation with one of MultiTag methods, check if the result is empty, then
 2. fallback to single tag estimation using a method like `estimateLowestAmbiguityPose`.
 

docs/source/docs/quick-start/quick-install.md

@@ -28,6 +28,10 @@ Unless otherwise noted in release notes or if updating from the prior years vers
 
 Use the [Raspberry Pi Imager](https://www.raspberrypi.com/software/) to flash the image onto the coprocessors microSD card. Select the downloaded `.img.xz` file, select your microSD card, and flash.
 
+:::{warning}
+Avoid using Raspberry Pi Imager version 2.0.2 or later. Those versions fail to write the image to an SD card. Versions 2.0.0 and earlier write images successfully. [GitHub issue 1489](https://github.com/raspberrypi/rpi-imager/issues/1489) was created for this problem.
+:::
+
 :::{warning}
 Balena Etcher has been recommended in the past, but should no longer be used due to instability and lack of ongoing support from developers.
 :::

docs/source/docs/quick-start/wiring.md

@@ -2,7 +2,11 @@
 
 ## Coprocessor with regulator
 
-1. **IT IS STRONGLY RECOMMENDED** to use one of the recommended power regulators to prevent vision from cutting out from voltage drops while operating the robot. We recommend wiring the regulator directly to the power header pins or using a locking USB C cable. In any case we recommend hot gluing the connector.
+1. **IT IS STRONGLY RECOMMENDED** to use one of the recommended power regulators to prevent vision from cutting out from voltage drops while operating the robot. We recommend wiring the regulator directly to the power header pins using either of the two methods listed below or using a locking USB C cable.
+    * Method 1: Soldering to GPIO Header Pins
+        * Using 20 AWG or preferably 18 AWG wires, solder two wires from the regulator to the power header pins on the coprocessor and cover with heat-shrink tubing.
+    * Method 2: Using a Wire-to-Board Connector
+        * Using a wire-to-board connector with 20 AWG or preferably 18 AWG wires, connect two wires from the regulator to the power header pins on the coprocessor. To prevent the connector from becoming unseated, we recommend applying hot glue to the connector.
 
 2. Run an ethernet cable from your coprocessor to your network switch / radio.
 

photon-client/src/components/app/photon-3d-visualizer.vue

@@ -44,7 +44,7 @@ let renderer: WebGLRenderer | undefined;
 let controls: TrackballControls | undefined;
 
 let previousTargets: Object3D[] = [];
-const drawTargets = (targets: PhotonTarget[]) => {
+const drawTargets = async (targets: PhotonTarget[]) => {
   // Check here, since if we check in watchEffect this never gets called
   if (!scene || !camera || !renderer || !controls) {
     return;
@@ -89,7 +89,11 @@ const drawTargets = (targets: PhotonTarget[]) => {
 
   if (calibrationCoeffs) {
     // And show camera frustum
-    const calibCamera = createPerspectiveCamera(calibrationCoeffs.resolution, calibrationCoeffs.cameraIntrinsics, 10);
+    const calibCamera = await createPerspectiveCamera(
+      calibrationCoeffs.resolution,
+      calibrationCoeffs.cameraIntrinsics,
+      10
+    );
     const helper = new CameraHelper(calibCamera);
     const helperGroup = new Group();
     helperGroup.add(helper);

photon-client/src/components/app/photon-calibration-visualizer.vue

@@ -65,7 +65,7 @@ const createChessboard = (obs: BoardObservation, cal: CameraCalibrationResult):
 
 let previousTargets: Object3D[] = [];
 let baseAspect: number | undefined;
-const drawCalibration = (cal: CameraCalibrationResult | null) => {
+const drawCalibration = async (cal: CameraCalibrationResult | null) => {
   // Check here, since if we check in watchEffect this never gets called
   if (!cal || !scene || !camera || !renderer || !controls) {
     return;
@@ -95,7 +95,7 @@ const drawCalibration = (cal: CameraCalibrationResult | null) => {
   });
 
   // And show camera frustum
-  const calibCamera = createPerspectiveCamera(props.resolution, cal.cameraIntrinsics);
+  const calibCamera = await createPerspectiveCamera(props.resolution, cal.cameraIntrinsics);
   const helper = new CameraHelper(calibCamera);
 
   // Flip to +Z forward

photon-client/src/components/cameras/CameraCalibrationCard.vue

@@ -14,6 +14,7 @@ import { getResolutionString, resolutionsAreEqual } from "@/lib/PhotonUtils";
 import CameraCalibrationInfoCard from "@/components/cameras/CameraCalibrationInfoCard.vue";
 import { useSettingsStore } from "@/stores/settings/GeneralSettingsStore";
 import { useTheme } from "vuetify";
+import TooltippedLabel from "@/components/common/pv-tooltipped-label.vue";
 
 const PromptRegular = import("@/assets/fonts/PromptRegular");
 const jspdf = import("jspdf");
@@ -99,6 +100,7 @@ const patternHeight = ref(8);
 const boardType = ref<CalibrationBoardTypes>(CalibrationBoardTypes.Charuco);
 const useOldPattern = ref(false);
 const tagFamily = ref<CalibrationTagFamilies>(CalibrationTagFamilies.Dict_4X4_1000);
+const requestedVideoFormatIndex = ref(0);
 
 // Emperical testing - with stack size limit of 1MB, we can handle at -least- 700k points
 const tooManyPoints = computed(
@@ -191,6 +193,7 @@ const startCalibration = () => {
   useCameraSettingsStore().currentCameraSettings.currentPipelineIndex = WebsocketPipelineType.Calib3d;
   // isCalibrating.value = true;
   calibCanceled.value = false;
+  requestedVideoFormatIndex.value = useStateStore().calibrationData.videoFormatIndex;
 };
 const showCalibEndDialog = ref(false);
 const calibCanceled = ref(false);
@@ -241,7 +244,14 @@ const setSelectedVideoFormat = (format: VideoFormat) => {
     <v-card class="mb-3 rounded-12" color="surface" dark>
       <v-card-title>Camera Calibration</v-card-title>
       <v-card-text v-if="!isCalibrating" class="pb-0">
-        <v-card-subtitle class="pa-0 pb-3 text-white">Current Calibrations</v-card-subtitle>
+        <div class="pb-3">
+          <tooltipped-label
+            label="Curent Calibrations"
+            icon="mdi-information"
+            location="top"
+            tooltip="Click on a resolution to view detailed calibration information and import/export a calibration."
+          />
+        </div>
         <v-table fixed-header height="100%" density="compact">
           <thead>
             <tr>
@@ -280,22 +290,10 @@ const setSelectedVideoFormat = (format: VideoFormat) => {
       </v-card-text>
       <v-card-text class="pt-0">
         <div v-if="useCameraSettingsStore().isConnected" class="d-flex flex-column">
-          <v-card-subtitle v-if="!isCalibrating" class="pl-0 pb-3 pt-3 text-white"
+          <v-card-subtitle v-if="!isCalibrating" class="pl-0 pb-3 pt-4 opacity-100"
             >Configure New Calibration</v-card-subtitle
           >
           <v-form ref="form" v-model="settingsValid">
-            <v-alert
-              closable
-              density="compact"
-              :variant="theme.global.name.value === 'LightTheme' ? 'elevated' : 'tonal'"
-              :color="useSettingsStore().general.mrCalWorking ? 'buttonPassive' : 'error'"
-              :icon="useSettingsStore().general.mrCalWorking ? 'mdi-check' : 'mdi-close'"
-              :text="
-                useSettingsStore().general.mrCalWorking
-                  ? 'Mrcal was successfully loaded and will be used!'
-                  : 'MrCal failed to load, check journalctl logs for details.'
-              "
-            />
             <pv-select
               v-model="uniqueVideoResolutionString"
               label="Resolution"
@@ -468,7 +466,20 @@ const setSelectedVideoFormat = (format: VideoFormat) => {
             "
           />
         </div>
-        <div v-if="isCalibrating" class="d-flex justify-center align-center pt-10px pb-5">
+        <v-alert
+          closable
+          density="compact"
+          class="mb-5"
+          :variant="theme.global.name.value === 'LightTheme' ? 'elevated' : 'tonal'"
+          :color="useSettingsStore().general.mrCalWorking ? 'buttonPassive' : 'error'"
+          :icon="useSettingsStore().general.mrCalWorking ? 'mdi-check' : 'mdi-close'"
+          :text="
+            useSettingsStore().general.mrCalWorking
+              ? 'Mrcal was successfully loaded and will be used!'
+              : 'MrCal failed to load, check journalctl logs for details.'
+          "
+        />
+        <div v-if="isCalibrating" class="d-flex justify-center align-center pb-5">
           <v-chip
             :variant="theme.global.name.value === 'LightTheme' ? 'elevated' : 'tonal'"
             label
@@ -559,7 +570,7 @@ const setSelectedVideoFormat = (format: VideoFormat) => {
               {{
                 useCameraSettingsStore().currentCameraSettings.validVideoFormats.map((f) =>
                   getResolutionString(f.resolution)
-                )[useStateStore().calibrationData.videoFormatIndex]
+                )[requestedVideoFormatIndex]
               }}!
             </v-card-text>
           </template>

photon-client/src/components/cameras/CameraCalibrationInfoCard.vue

@@ -187,7 +187,7 @@ const viewingImg = ref(0);
                           .getCalibrationCoeffs(props.videoFormat.resolution)
                           ?.cameraIntrinsics.data[0].toFixed(2) || 0.0
                       }}
-                      mm
+                      px
                     </td>
                   </tr>
                   <tr>
@@ -198,7 +198,7 @@ const viewingImg = ref(0);
                           .getCalibrationCoeffs(props.videoFormat.resolution)
                           ?.cameraIntrinsics.data[4].toFixed(2) || 0.0
                       }}
-                      mm
+                      px
                     </td>
                   </tr>
                   <tr>

photon-client/src/components/dashboard/tabs/OutputTab.vue

@@ -3,6 +3,7 @@ import PvSelect from "@/components/common/pv-select.vue";
 import { useCameraSettingsStore } from "@/stores/settings/CameraSettingsStore";
 import { type ActivePipelineSettings, PipelineType, RobotOffsetPointMode } from "@/types/PipelineTypes";
 import PvSwitch from "@/components/common/pv-switch.vue";
+import PvSlider from "@/components/common/pv-slider.vue";
 import { computed } from "vue";
 import { RobotOffsetType } from "@/types/SettingTypes";
 import { useStateStore } from "@/stores/StateStore";
@@ -58,14 +59,17 @@ const interactiveCols = computed(() =>
 
 <template>
   <div>
-    <pv-switch
-      v-model="useCameraSettingsStore().currentPipelineSettings.outputShowMultipleTargets"
-      label="Show Multiple Targets"
-      tooltip="If enabled, up to five targets will be displayed and sent via PhotonLib, instead of just one"
-      :disabled="isTagPipeline"
+    <pv-slider
+      v-model="useCameraSettingsStore().currentPipelineSettings.outputMaximumTargets"
+      label="Maximum Targets"
+      tooltip="The maximum number of targets to display and send."
+      :hidden="isTagPipeline"
+      :min="1"
+      :max="127"
+      :step="1"
       :switch-cols="interactiveCols"
       @update:modelValue="
-        (value) => useCameraSettingsStore().changeCurrentPipelineSetting({ outputShowMultipleTargets: value }, false)
+        (value) => useCameraSettingsStore().changeCurrentPipelineSetting({ outputMaximumTargets: value }, false)
       "
     />
     <pv-switch

photon-client/src/components/settings/DeviceCard.vue

@@ -13,13 +13,15 @@ import { metricsHistorySnapshot } from "@/stores/settings/GeneralSettingsStore";
 
 const theme = useTheme();
 
-const restartProgram = () => {
-  axiosPost("/utils/restartProgram", "restart PhotonVision");
-  forceReloadPage();
+const restartProgram = async () => {
+  if (await axiosPost("/utils/restartProgram", "restart PhotonVision")) {
+    forceReloadPage();
+  }
 };
-const restartDevice = () => {
-  axiosPost("/utils/restartDevice", "restart the device");
-  forceReloadPage();
+const restartDevice = async () => {
+  if (await axiosPost("/utils/restartDevice", "restart the device")) {
+    forceReloadPage();
+  }
 };
 
 const address = inject<string>("backendHost");
@@ -38,28 +40,30 @@ const handleOfflineUpdate = async () => {
     color: "secondary",
     timeout: -1
   });
-  await axiosPost("/utils/offlineUpdate", "upload new software", formData, {
-    headers: { "Content-Type": "multipart/form-data" },
-    onUploadProgress: ({ progress }) => {
-      const uploadPercentage = (progress || 0) * 100.0;
-      if (uploadPercentage < 99.5) {
-        useStateStore().showSnackbarMessage({
-          message: "New Software Upload in Progress",
-          color: "secondary",
-          timeout: -1,
-          progressBar: uploadPercentage,
-          progressBarColor: "primary"
-        });
-      } else {
-        useStateStore().showSnackbarMessage({
-          message: "Installing uploaded software...",
-          color: "secondary",
-          timeout: -1
-        });
+  if (
+    await axiosPost("/utils/offlineUpdate", "upload new software", formData, {
+      headers: { "Content-Type": "multipart/form-data" },
+      onUploadProgress: ({ progress }) => {
+        const uploadPercentage = (progress || 0) * 100.0;
+        if (uploadPercentage < 99.5) {
+          useStateStore().showSnackbarMessage({
+            message: "New Software Upload in Progress",
+            color: "secondary",
+            timeout: -1,
+            progressBar: uploadPercentage,
+            progressBarColor: "primary"
+          });
+        }
       }
-    }
-  });
-  forceReloadPage();
+    })
+  ) {
+    useStateStore().showSnackbarMessage({
+      message: "Installing uploaded software...",
+      color: "secondary",
+      timeout: -1
+    });
+    forceReloadPage();
+  }
 };
 
 const exportLogFile = ref();
@@ -116,9 +120,10 @@ const handleSettingsImport = () => {
 };
 
 const showFactoryReset = ref(false);
-const nukePhotonConfigDirectory = () => {
-  axiosPost("/utils/nukeConfigDirectory", "delete the config directory");
-  forceReloadPage();
+const nukePhotonConfigDirectory = async () => {
+  if (await axiosPost("/utils/nukeConfigDirectory", "delete the config directory")) {
+    forceReloadPage();
+  }
 };
 
 interface MetricItem {
@@ -371,21 +376,33 @@ watch(metricsHistorySnapshot, () => {
             <span>CPU Usage</span>
             <span>{{ Math.round(cpuUsageData.at(-1)?.value ?? 0) }}%</span>
           </div>
-          <MetricsChart id="chart" :data="cpuUsageData" type="percentage" :min="0" :max="100" color="blue" />
+          <Suspense>
+            <!-- Allows us to import echarts when it's actually needed  -->
+            <MetricsChart id="chart" :data="cpuUsageData" type="percentage" :min="0" :max="100" color="blue" />
+            <template #fallback> Loading... </template>
+          </Suspense>
         </v-card-text>
         <v-card-text class="pt-0 flex-0-0 pb-2">
           <div class="d-flex justify-space-between pb-3 pt-3">
             <span>CPU Memory Usage</span>
             <span>{{ Math.round(cpuMemoryUsageData.at(-1)?.value ?? 0) }}%</span>
           </div>
-          <MetricsChart id="chart" :data="cpuMemoryUsageData" type="percentage" :min="0" :max="100" color="purple" />
+          <Suspense>
+            <!-- Allows us to import echarts when it's actually needed  -->
+            <MetricsChart id="chart" :data="cpuMemoryUsageData" type="percentage" :min="0" :max="100" color="purple" />
+            <template #fallback> Loading... </template>
+          </Suspense>
         </v-card-text>
         <v-card-text class="pt-0 flex-0-0 pb-2">
           <div class="d-flex justify-space-between pb-3 pt-3">
             <span>CPU Temperature</span>
             <span>{{ cpuTempData.at(-1)?.value == -1 ? "--- " : Math.round(cpuTempData.at(-1)?.value ?? 0) }}°C</span>
           </div>
-          <MetricsChart id="chart" :data="cpuTempData" type="temperature" color="red" />
+          <Suspense>
+            <!-- Allows us to import echarts when it's actually needed  -->
+            <MetricsChart id="chart" :data="cpuTempData" type="temperature" color="red" />
+            <template #fallback> Loading... </template>
+          </Suspense>
         </v-card-text>
         <v-card-text class="pt-0 flex-0-0">
           <div class="d-flex justify-space-between pb-3 pt-3">
@@ -399,7 +416,11 @@ watch(metricsHistorySnapshot, () => {
               >{{ networkUsageData.at(-1)?.value == -1 ? "---" : networkUsageData.at(-1)?.value.toFixed(3) }} Mb/s</span
             >
           </div>
-          <MetricsChart id="chart" :data="networkUsageData" type="mb" :min="0" :max="10" color="green" />
+          <Suspense>
+            <!-- Allows us to import echarts when it's actually needed  -->
+            <MetricsChart id="chart" :data="networkUsageData" type="mb" :min="0" :max="10" color="green" />
+            <template #fallback> Loading... </template>
+          </Suspense>
         </v-card-text>
       </v-card>
     </v-col>

photon-client/src/components/settings/MetricsChart.vue

@@ -1,5 +1,4 @@
 <script setup lang="ts">
-import * as echarts from "echarts";
 import { onMounted, ref, onBeforeUnmount, watch } from "vue";
 import { useTheme } from "vuetify";
 
@@ -198,7 +197,8 @@ const props = defineProps<{
   color?: string;
 }>();
 
-onMounted(() => {
+onMounted(async () => {
+  const echarts = await import("echarts");
   chart = echarts.init(chartRef.value);
   chart.setOption(getOptions(props.data));
 

photon-client/src/components/settings/ObjectDetectionCard.vue

@@ -43,25 +43,27 @@ const handleImport = async () => {
     timeout: -1
   });
 
-  axiosPost("/objectdetection/import", "import an object detection model", formData, {
-    headers: { "Content-Type": "multipart/form-data" },
-    onUploadProgress: ({ progress }) => {
-      const uploadPercentage = (progress || 0) * 100.0;
-      if (uploadPercentage < 99.5) {
-        useStateStore().showSnackbarMessage({
-          message: "Object Detection Model Upload in Process, " + uploadPercentage.toFixed(2) + "% complete",
-          color: "secondary",
-          timeout: -1
-        });
-      } else {
-        useStateStore().showSnackbarMessage({
-          message: "Processing uploaded Object Detection Model...",
-          color: "secondary",
-          timeout: -1
-        });
+  if (
+    await axiosPost("/objectdetection/import", "import an object detection model", formData, {
+      headers: { "Content-Type": "multipart/form-data" },
+      onUploadProgress: ({ progress }) => {
+        const uploadPercentage = (progress || 0) * 100.0;
+        if (uploadPercentage < 99.5) {
+          useStateStore().showSnackbarMessage({
+            message: "Object Detection Model Upload in Process, " + uploadPercentage.toFixed(2) + "% complete",
+            color: "secondary",
+            timeout: -1
+          });
+        }
       }
-    }
-  });
+    })
+  ) {
+    useStateStore().showSnackbarMessage({
+      message: "Processing uploaded Object Detection Model...",
+      color: "secondary",
+      timeout: -1
+    });
+  }
 
   showImportDialog.value = false;
 
@@ -72,13 +74,13 @@ const handleImport = async () => {
   importVersion.value = null;
 };
 
-const deleteModel = async (model: ObjectDetectionModelProperties) => {
+const deleteModel = (model: ObjectDetectionModelProperties) => {
   axiosPost("/objectdetection/delete", "delete an object detection model", {
     modelPath: model.modelPath
   });
 };
 
-const renameModel = async (model: ObjectDetectionModelProperties, newName: string) => {
+const renameModel = (model: ObjectDetectionModelProperties, newName: string) => {
   useStateStore().showSnackbarMessage({
     message: "Renaming Object Detection Model...",
     color: "secondary",
@@ -121,33 +123,35 @@ const nukeModels = () => {
 
 const showBulkImportDialog = ref(false);
 const importFile = ref<File | null>(null);
-const handleBulkImport = () => {
+const handleBulkImport = async () => {
   if (importFile.value === null) return;
 
   const formData = new FormData();
   formData.append("data", importFile.value);
 
-  axiosPost("/objectdetection/bulkimport", "import object detection models", formData, {
-    headers: { "Content-Type": "multipart/form-data" },
-    onUploadProgress: ({ progress }) => {
-      const uploadPercentage = (progress || 0) * 100.0;
-      if (uploadPercentage < 99.5) {
-        useStateStore().showSnackbarMessage({
-          message: "Object Detection Models Upload in Progress",
-          color: "secondary",
-          timeout: -1,
-          progressBar: uploadPercentage,
-          progressBarColor: "primary"
-        });
-      } else {
-        useStateStore().showSnackbarMessage({
-          message: "Importing New Object Detection Models...",
-          color: "secondary",
-          timeout: -1
-        });
+  if (
+    await axiosPost("/objectdetection/bulkimport", "import object detection models", formData, {
+      headers: { "Content-Type": "multipart/form-data" },
+      onUploadProgress: ({ progress }) => {
+        const uploadPercentage = (progress || 0) * 100.0;
+        if (uploadPercentage < 99.5) {
+          useStateStore().showSnackbarMessage({
+            message: "Object Detection Models Upload in Progress",
+            color: "secondary",
+            timeout: -1,
+            progressBar: uploadPercentage,
+            progressBarColor: "primary"
+          });
+        }
       }
-    }
-  });
+    })
+  ) {
+    useStateStore().showSnackbarMessage({
+      message: "Importing New Object Detection Models...",
+      color: "secondary",
+      timeout: -1
+    });
+  }
   showImportDialog.value = false;
   importFile.value = null;
 };

photon-client/src/lib/PhotonUtils.ts

@@ -22,7 +22,7 @@ export const statusCheck = async (timeout: number, ip?: string): Promise<boolean
   while (pollLimit > 0) {
     try {
       pollLimit--;
-      await axios.get(ip ? `http://${ip}/status` : "/status");
+      await axios.get(ip ? `http://${ip}/api/status` : "/status");
       return true;
     } catch {
       // Backend not ready yet, wait and retry
@@ -71,33 +71,33 @@ export const parseJsonFile = async <T extends Record<string, any>>(file: File):
  * @param description A brief description of the request for users, e.g., "import object detection models".
  * @param data Payload to be sent in the POST request
  * @param config Optional axios request configuration
- * @returns A promise that resolves when the POST request is complete
+ * @returns A promise that resolves to true if the POST request is successful, or false if an error occurs.
  */
-export const axiosPost = (url: string, description: string, data?: any, config?: any): Promise<void> => {
-  return axios
-    .post(url, data, config)
-    .then(() => {
-      useStateStore().showSnackbarMessage({
-        message: "Successfully dispatched the request to " + description + ". Waiting for backend to respond",
-        color: "success"
-      });
-    })
-    .catch((error) => {
-      if (error.response) {
-        useStateStore().showSnackbarMessage({
-          message: "The backend is unable to fulfill the request to " + description + ".",
-          color: "error"
-        });
-      } else if (error.request) {
-        useStateStore().showSnackbarMessage({
-          message: "Error while trying to process the request to " + description + "! The backend didn't respond.",
-          color: "error"
-        });
-      } else {
-        useStateStore().showSnackbarMessage({
-          message: "An error occurred while trying to process the request to " + description + ".",
-          color: "error"
-        });
-      }
+export const axiosPost = async (url: string, description: string, data?: any, config?: any): Promise<boolean> => {
+  try {
+    await axios.post(url, data, config);
+    useStateStore().showSnackbarMessage({
+      message: "Successfully dispatched the request to " + description + ". Waiting for backend to respond",
+      color: "success"
     });
+    return true;
+  } catch (error: any) {
+    if (error.response) {
+      useStateStore().showSnackbarMessage({
+        message: "The backend is unable to fulfill the request to " + description + ".",
+        color: "error"
+      });
+    } else if (error.request) {
+      useStateStore().showSnackbarMessage({
+        message: "Error while trying to process the request to " + description + "! The backend didn't respond.",
+        color: "error"
+      });
+    } else {
+      useStateStore().showSnackbarMessage({
+        message: "An error occurred while trying to process the request to " + description + ".",
+        color: "error"
+      });
+    }
+    return false;
+  }
 };

photon-client/src/lib/ThreeUtils.ts

@@ -1,5 +1,5 @@
 import type { JsonMatOfDouble, Resolution } from "@/types/SettingTypes";
-const { PerspectiveCamera } = await import("three");
+const three = import("three");
 
 /**
  * Convert a camera intrinsics matrix and image resolution to a Three.js PerspectiveCamera. This assumes no skew and square pixels (same focal length in x and y), which is a sane assumption for most FRC cameras
@@ -8,11 +8,12 @@ const { PerspectiveCamera } = await import("three");
  * @param intrinsicsCore camera intrinsics from the backend, row-major
  * @returns a Three.js PerspectiveCamera matching the provided intrinsics
  */
-export const createPerspectiveCamera = (
+export const createPerspectiveCamera = async (
   resolution: Resolution,
   intrinsicsCore: JsonMatOfDouble,
   frustumMax: number = 1
 ) => {
+  const { PerspectiveCamera } = await three;
   const imageWidth = resolution.width;
   const imageHeight = resolution.height;
   const focalLengthY = intrinsicsCore.data[4];

photon-client/src/types/PipelineTypes.ts

@@ -66,7 +66,7 @@ export interface PipelineSettings {
   hsvHue: WebsocketNumberPair | [number, number];
   ledMode: boolean;
   hueInverted: boolean;
-  outputShowMultipleTargets: boolean;
+  outputMaximumTargets: number;
   contourSortMode: number;
   cameraExposureRaw: number;
   cameraMinExposureRaw: number;
@@ -108,7 +108,7 @@ export type ConfigurablePipelineSettings = Partial<
 // Omitted settings are changed for all pipeline types
 export const DefaultPipelineSettings: Omit<
   PipelineSettings,
-  "cameraGain" | "targetModel" | "ledMode" | "outputShowMultipleTargets" | "cameraExposureRaw" | "pipelineType"
+  "cameraGain" | "targetModel" | "ledMode" | "cameraExposureRaw" | "pipelineType"
 > = {
   offsetRobotOffsetMode: RobotOffsetPointMode.None,
   streamingFrameDivisor: 0,
@@ -137,6 +137,7 @@ export const DefaultPipelineSettings: Omit<
   offsetDualPointB: { x: 0, y: 0 },
   hsvHue: { first: 50, second: 180 },
   hueInverted: false,
+  outputMaximumTargets: 20,
   contourSortMode: 0,
   offsetSinglePoint: { x: 0, y: 0 },
   cameraBrightness: 50,
@@ -166,7 +167,7 @@ export const DefaultReflectivePipelineSettings: ReflectivePipelineSettings = {
   cameraGain: 20,
   targetModel: TargetModel.InfiniteRechargeHighGoalOuter,
   ledMode: true,
-  outputShowMultipleTargets: false,
+  outputMaximumTargets: 20,
   cameraExposureRaw: 6,
   pipelineType: PipelineType.Reflective,
 
@@ -197,7 +198,7 @@ export const DefaultColoredShapePipelineSettings: ColoredShapePipelineSettings =
   cameraGain: 75,
   targetModel: TargetModel.InfiniteRechargeHighGoalOuter,
   ledMode: true,
-  outputShowMultipleTargets: false,
+  outputMaximumTargets: 20,
   cameraExposureRaw: 20,
   pipelineType: PipelineType.ColoredShape,
 
@@ -237,10 +238,9 @@ export const DefaultAprilTagPipelineSettings: AprilTagPipelineSettings = {
   cameraGain: 75,
   targetModel: TargetModel.AprilTag6p5in_36h11,
   ledMode: false,
-  outputShowMultipleTargets: true,
+  outputMaximumTargets: 127,
   cameraExposureRaw: 20,
   pipelineType: PipelineType.AprilTag,
-
   hammingDist: 0,
   numIterations: 40,
   decimate: 1,
@@ -278,13 +278,12 @@ export type ConfigurableArucoPipelineSettings = Partial<Omit<ArucoPipelineSettin
 export const DefaultArucoPipelineSettings: ArucoPipelineSettings = {
   ...DefaultPipelineSettings,
   cameraGain: 75,
-  outputShowMultipleTargets: true,
+  outputMaximumTargets: 127,
   targetModel: TargetModel.AprilTag6p5in_36h11,
   cameraExposureRaw: -1,
   cameraAutoExposure: true,
   ledMode: false,
   pipelineType: PipelineType.Aruco,
-
   tagFamily: AprilTagFamily.Family36h11,
   threshWinSizes: { first: 11, second: 91 },
   threshStepSize: 40,
@@ -316,7 +315,7 @@ export const DefaultObjectDetectionPipelineSettings: ObjectDetectionPipelineSett
   cameraGain: 20,
   targetModel: TargetModel.InfiniteRechargeHighGoalOuter,
   ledMode: true,
-  outputShowMultipleTargets: false,
+  outputMaximumTargets: 20,
   cameraExposureRaw: 6,
   confidence: 0.9,
   nms: 0.45,
@@ -335,7 +334,7 @@ export const DefaultCalibration3dPipelineSettings: Calibration3dPipelineSettings
   cameraGain: 20,
   targetModel: TargetModel.InfiniteRechargeHighGoalOuter,
   ledMode: true,
-  outputShowMultipleTargets: false,
+  outputMaximumTargets: 1,
   cameraExposureRaw: 6,
   drawAllSnapshots: false
 };

photon-core/src/main/java/org/photonvision/common/configuration/ConfigManager.java

@@ -39,7 +39,7 @@ import org.photonvision.vision.processes.VisionSource;
 import org.zeroturnaround.zip.ZipUtil;
 
 public class ConfigManager {
-    private static ConfigManager INSTANCE;
+    static ConfigManager INSTANCE;
 
     public static final String HW_CFG_FNAME = "hardwareConfig.json";
     public static final String HW_SET_FNAME = "hardwareSettings.json";

photon-core/src/main/java/org/photonvision/common/configuration/LegacyConfigProvider.java

@@ -218,7 +218,7 @@ class LegacyConfigProvider extends ConfigProvider {
                         hardwareSettings,
                         networkConfig,
                         atfl,
-                        new NeuralNetworkPropertyManager(),
+                        new NeuralNetworkModelsSettings(),
                         cameraConfigurations);
     }
 

photon-core/src/main/java/org/photonvision/common/configuration/NeuralNetworkModelManager.java

@@ -34,7 +34,7 @@ import java.util.Optional;
 import java.util.jar.JarEntry;
 import java.util.jar.JarFile;
 import java.util.stream.Stream;
-import org.photonvision.common.configuration.NeuralNetworkPropertyManager.ModelProperties;
+import org.photonvision.common.configuration.NeuralNetworkModelsSettings.ModelProperties;
 import org.photonvision.common.hardware.Platform;
 import org.photonvision.common.logging.LogGroup;
 import org.photonvision.common.logging.Logger;
@@ -50,20 +50,20 @@ import org.photonvision.vision.objects.RubikModel;
  * extracted to the filesystem, it will not be extracted again.
  *
  * <p>Each model must have a corresponding {@link ModelProperties} entry in {@link
- * NeuralNetworkPropertyManager}.
+ * NeuralNetworkModelsSettings}.
  */
 public class NeuralNetworkModelManager {
     /** Singleton instance of the NeuralNetworkModelManager */
     private static NeuralNetworkModelManager INSTANCE;
 
-    private final List<Family> supportedBackends = new ArrayList<>();
+    final List<Family> supportedBackends = new ArrayList<>();
 
     /**
      * This function stores the properties of the shipped object detection models. It is stored as a
      * function so that it can be dynamic, to adjust for the models directory.
      */
-    private NeuralNetworkPropertyManager getShippedProperties(File modelsDirectory) {
-        NeuralNetworkPropertyManager nnProps = new NeuralNetworkPropertyManager();
+    private NeuralNetworkModelsSettings getShippedProperties(File modelsDirectory) {
+        NeuralNetworkModelsSettings nnProps = new NeuralNetworkModelsSettings();
 
         LinkedList<String> cocoLabels =
                 new LinkedList<String>(
@@ -146,19 +146,9 @@ public class NeuralNetworkModelManager {
                                 "vase",
                                 "scissors",
                                 "teddy bear",
-                                "hair drier",
+                                "hair drier", // Typo in official COCO documentation
                                 "toothbrush"));
 
-        nnProps.addModelProperties(
-                new ModelProperties(
-                        Path.of(modelsDirectory.getAbsolutePath(), "algaeV1-640-640-yolov8n.rknn"),
-                        "Algae v8n",
-                        new LinkedList<String>(List.of("Algae")),
-                        640,
-                        480,
-                        Family.RKNN,
-                        Version.YOLOV8));
-
         nnProps.addModelProperties(
                 new ModelProperties(
                         Path.of(modelsDirectory.getAbsolutePath(), "yolov8nCOCO.rknn"),
@@ -171,13 +161,13 @@ public class NeuralNetworkModelManager {
 
         nnProps.addModelProperties(
                 new ModelProperties(
-                        Path.of(modelsDirectory.getAbsolutePath(), "algae-coral-yolov8s.tflite"),
-                        "Algae Coral v8s",
-                        new LinkedList<String>(List.of("Algae", "Coral")),
+                        Path.of(modelsDirectory.getAbsolutePath(), "fuelV1-yolo11n.rknn"),
+                        "Fuel v11n",
+                        new LinkedList<String>(List.of("Fuel")),
                         640,
                         640,
-                        Family.RUBIK,
-                        Version.YOLOV8));
+                        Family.RKNN,
+                        Version.YOLOV11));
 
         nnProps.addModelProperties(
                 new ModelProperties(
@@ -189,6 +179,16 @@ public class NeuralNetworkModelManager {
                         Family.RUBIK,
                         Version.YOLOV8));
 
+        nnProps.addModelProperties(
+                new ModelProperties(
+                        Path.of(modelsDirectory.getAbsolutePath(), "fuelV1-yolo11n.tflite"),
+                        "Fuel v11n",
+                        new LinkedList<String>(List.of("Fuel")),
+                        640,
+                        640,
+                        Family.RUBIK,
+                        Version.YOLOV11));
+
         return nnProps;
     }
 
@@ -272,7 +272,7 @@ public class NeuralNetworkModelManager {
      *
      * <p>The first model in the list is the default model.
      */
-    private Map<Family, ArrayList<Model>> models;
+    Map<Family, ArrayList<Model>> models;
 
     /**
      * Retrieves the model with the specified name, assuming it is available under a supported
@@ -321,13 +321,27 @@ public class NeuralNetworkModelManager {
                 ConfigManager.getInstance().getConfig().neuralNetworkPropertyManager().getModel(path);
 
         if (properties == null) {
-            logger.error(
+            logger.warn(
                     "Model properties are null. This could mean the config for model "
                             + path
-                            + " was unable to be found in the database.");
-            return;
+                            + " was unable to be found in the database. Trying legacy...");
+            try {
+                properties = ModelProperties.createFromFilename(path.getFileName().toString());
+
+                // At this point this property is not serialized or known to our configuration. add to
+                // NeuralNetworkModelsSettings
+                ConfigManager.getInstance()
+                        .getConfig()
+                        .neuralNetworkPropertyManager()
+                        .addModelProperties(properties);
+            } catch (IllegalArgumentException | IOException e) {
+                logger.error("Failed to translate legacy model filename to properties: " + path, e);
+                return;
+            }
         }
 
+        logger.debug(properties.toString());
+
         if (!supportedBackends.contains(properties.family())) {
             logger.warn(
                     "Model "
@@ -412,7 +426,7 @@ public class NeuralNetworkModelManager {
         File modelsDirectory = ConfigManager.getInstance().getModelsDirectory();
 
         // Filter shippedProprties by supportedBackends
-        NeuralNetworkPropertyManager supportedProperties = new NeuralNetworkPropertyManager();
+        NeuralNetworkModelsSettings supportedProperties = new NeuralNetworkModelsSettings();
         for (ModelProperties model : getShippedProperties(modelsDirectory).getModels()) {
             if (supportedBackends.contains(model.family())) {
                 supportedProperties.addModelProperties(model);

photon-core/src/main/java/org/photonvision/common/configuration/NeuralNetworkModelsSettings.java

@@ -0,0 +1,277 @@
+/*
+ * Copyright (C) Photon Vision.
+ *
+ * This program is free software: you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation, either version 3 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <https://www.gnu.org/licenses/>.
+ */
+
+package org.photonvision.common.configuration;
+
+import com.fasterxml.jackson.annotation.JsonCreator;
+import com.fasterxml.jackson.annotation.JsonIgnore;
+import com.fasterxml.jackson.annotation.JsonProperty;
+import java.io.IOException;
+import java.nio.file.Files;
+import java.nio.file.Path;
+import java.nio.file.Paths;
+import java.util.Arrays;
+import java.util.HashMap;
+import java.util.List;
+import java.util.regex.Matcher;
+import java.util.regex.Pattern;
+import org.photonvision.common.configuration.NeuralNetworkModelManager.Family;
+import org.photonvision.common.configuration.NeuralNetworkModelManager.Version;
+
+public class NeuralNetworkModelsSettings {
+    /*
+     * The properties of the model. This is used to determine which model to load.
+     * The only families currently supported are RKNN and Rubik (custom .tflite)
+     */
+    public record ModelProperties(
+            @JsonProperty("modelPath") Path modelPath,
+            @JsonProperty("nickname") String nickname,
+            @JsonProperty("labels") List<String> labels,
+            @JsonProperty("resolutionWidth") int resolutionWidth,
+            @JsonProperty("resolutionHeight") int resolutionHeight,
+            @JsonProperty("family") Family family,
+            @JsonProperty("version") Version version) {
+        @JsonCreator
+        public ModelProperties {}
+
+        ModelProperties(ModelProperties other) {
+            this(
+                    other.modelPath,
+                    other.nickname,
+                    other.labels, // note this does not clone the underlying list
+                    other.resolutionWidth,
+                    other.resolutionHeight,
+                    other.family,
+                    other.version);
+        }
+
+        // In v2025.3.1, this was single string for the model path. but the first argument
+        // is now nickname
+        public ModelProperties(@JsonProperty("nickname") String filename)
+                throws IllegalArgumentException, IOException {
+            this(createFromFilename(filename));
+        }
+
+        // ============= Migration code from v2025.3.1 ===========
+
+        private static Pattern modelPattern =
+                Pattern.compile("^([a-zA-Z0-9._]+)-(\\d+)-(\\d+)-(yolov(?:5|8|11)[nsmlx]*)\\.rknn$");
+
+        static ModelProperties createFromFilename(String modelFileName)
+                throws IllegalArgumentException, IOException {
+            // Used to point to default models directory
+            var model =
+                    ConfigManager.getInstance().getModelsDirectory().toPath().resolve(modelFileName).toFile();
+
+            // Get the model extension and check if it is supported
+            String modelExtension = model.getName().substring(model.getName().lastIndexOf('.'));
+            if (!modelExtension.equals(".rknn")) {
+                throw new IllegalArgumentException("Model " + modelFileName + " is not a supported format");
+            }
+
+            var backend =
+                    Arrays.stream(NeuralNetworkModelManager.Family.values())
+                            .filter(b -> b.extension().equals(modelExtension))
+                            .findFirst();
+
+            if (!backend.isPresent()) {
+                throw new IllegalArgumentException("Model " + modelFileName + " cannot find backend");
+            }
+
+            String labelFile = model.getAbsolutePath().replace(backend.get().extension(), "-labels.txt");
+            List<String> labels = Files.readAllLines(Paths.get(labelFile));
+
+            String[] parts = parseRKNNName(modelFileName);
+            var version = getModelVersion(parts[3]);
+            int width = Integer.parseInt(parts[1]);
+            int height = Integer.parseInt(parts[2]);
+
+            return new ModelProperties(
+                    model.toPath(),
+                    model.getName(),
+                    labels,
+                    // all files used to be 640x640
+                    width,
+                    height,
+                    Family.RKNN,
+                    version);
+        }
+
+        /**
+         * Determines the model version based on the model's filename.
+         *
+         * <p>"yolov5" -> "YOLO_V5"
+         *
+         * <p>"yolov8" -> "YOLO_V8"
+         *
+         * <p>"yolov11" -> "YOLO_V11"
+         *
+         * @param modelName The model's filename
+         * @return The model version
+         */
+        private static Version getModelVersion(String modelName) throws IllegalArgumentException {
+            if (modelName.contains("yolov5")) {
+                return Version.YOLOV5;
+            } else if (modelName.contains("yolov8")) {
+                return Version.YOLOV8;
+            } else if (modelName.contains("yolov11")) {
+                return Version.YOLOV11;
+            } else {
+                throw new IllegalArgumentException("Unknown model version for model " + modelName);
+            }
+        }
+
+        /**
+         * Parse RKNN name and return the name, width, height, and model type.
+         *
+         * <p>This is static as it is not dependent on the state of the class.
+         *
+         * @param modelName the name of the model
+         * @throws IllegalArgumentException if the model name does not follow the naming convention
+         * @return an array containing the name, width, height, and model type
+         */
+        public static String[] parseRKNNName(String modelName) {
+            Matcher modelMatcher = modelPattern.matcher(modelName);
+
+            if (!modelMatcher.matches()) {
+                throw new IllegalArgumentException(
+                        "Model name must follow the naming convention of name-widthResolution-heightResolution-modelType.rknn");
+            }
+
+            return new String[] {
+                modelMatcher.group(1), modelMatcher.group(2), modelMatcher.group(3), modelMatcher.group(4)
+            };
+        }
+    }
+
+    // The path to the model is used as the key in the map because it is unique to
+    // the model, and should not change
+    @JsonProperty("modelPathToProperties")
+    private HashMap<Path, ModelProperties> modelPathToProperties =
+            new HashMap<Path, ModelProperties>();
+
+    /**
+     * Constructor for the NeuralNetworkProperties class.
+     *
+     * <p>This object holds a LinkedList of {@link ModelProperties} objects
+     */
+    public NeuralNetworkModelsSettings() {}
+
+    /**
+     * Constructor for the NeuralNetworkProperties class.
+     *
+     * <p>This object holds a LinkedList of {@link ModelProperties} objects.
+     *
+     * @param modelPropertiesList When the class is constructed, it will hold the provided list
+     */
+    public NeuralNetworkModelsSettings(HashMap<Path, ModelProperties> modelPropertiesList) {}
+
+    @Override
+    public String toString() {
+        String toReturn = "";
+
+        toReturn += "NeuralNetworkProperties [";
+
+        toReturn += modelPathToProperties.toString() + "]";
+
+        return toReturn;
+    }
+
+    /**
+     * Add a model to the list of models.
+     *
+     * @param modelProperties
+     */
+    public void addModelProperties(ModelProperties modelProperties) {
+        modelPathToProperties.put(modelProperties.modelPath, modelProperties);
+    }
+
+    /**
+     * Add two Neural Network Properties together.
+     *
+     * <p>Any properties that are the same will be overwritten by the second
+     *
+     * @param nnProps
+     * @return itself, so it can be chained and used fluently
+     */
+    public NeuralNetworkModelsSettings sum(NeuralNetworkModelsSettings nnProps) {
+        modelPathToProperties.putAll(nnProps.modelPathToProperties);
+
+        return this;
+    }
+
+    /**
+     * Remove a model from the list of models.
+     *
+     * @param modelPath
+     * @return True if the model was removed, false if it was not found
+     */
+    public boolean removeModel(Path modelPath) {
+        return modelPathToProperties.remove(modelPath) != null;
+    }
+
+    /**
+     * Get the model properties for a given model path.
+     *
+     * @param modelPath
+     * @return {@link ModelProperties} object
+     */
+    public ModelProperties getModel(Path modelPath) {
+        return modelPathToProperties.get(modelPath);
+    }
+
+    /**
+     * Get all models
+     *
+     * @return A list of all models
+     */
+    @JsonIgnore
+    public ModelProperties[] getModels() {
+        return modelPathToProperties.values().toArray(new ModelProperties[0]);
+    }
+
+    /**
+     * Change the nickname of a {@link ModelProperties} object.
+     *
+     * @param modelPath
+     * @param newName
+     * @return True if the model was found and renamed, false if it was not found
+     */
+    public boolean renameModel(Path modelPath, String newName) {
+        ModelProperties temp = modelPathToProperties.get(modelPath);
+        if (temp != null) {
+            modelPathToProperties.remove(modelPath);
+            modelPathToProperties.put(
+                    modelPath,
+                    new ModelProperties(
+                            temp.modelPath,
+                            newName,
+                            temp.labels,
+                            temp.resolutionWidth,
+                            temp.resolutionHeight,
+                            temp.family,
+                            temp.version));
+            return true;
+        }
+        return false;
+    }
+
+    public boolean clear() {
+        modelPathToProperties.clear();
+        return true;
+    }
+}

photon-core/src/main/java/org/photonvision/common/configuration/NeuralNetworkPropertyManager.java

@@ -1,164 +0,0 @@
-/*
- * Copyright (C) Photon Vision.
- *
- * This program is free software: you can redistribute it and/or modify
- * it under the terms of the GNU General Public License as published by
- * the Free Software Foundation, either version 3 of the License, or
- * (at your option) any later version.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program.  If not, see <https://www.gnu.org/licenses/>.
- */
-
-package org.photonvision.common.configuration;
-
-import com.fasterxml.jackson.annotation.JsonCreator;
-import com.fasterxml.jackson.annotation.JsonIgnore;
-import com.fasterxml.jackson.annotation.JsonProperty;
-import java.nio.file.Path;
-import java.util.HashMap;
-import java.util.LinkedList;
-import org.photonvision.common.configuration.NeuralNetworkModelManager.Family;
-import org.photonvision.common.configuration.NeuralNetworkModelManager.Version;
-
-public class NeuralNetworkPropertyManager {
-    /*
-     * The properties of the model. This is used to determine which model to load.
-     * The only families currently supported are RKNN and Rubik (custom .tflite)
-     */
-    public record ModelProperties(
-            @JsonProperty("modelPath") Path modelPath,
-            @JsonProperty("nickname") String nickname,
-            @JsonProperty("labels") LinkedList<String> labels,
-            @JsonProperty("resolutionWidth") int resolutionWidth,
-            @JsonProperty("resolutionHeight") int resolutionHeight,
-            @JsonProperty("family") Family family,
-            @JsonProperty("version") Version version) {
-        @JsonCreator
-        public ModelProperties {
-            // Record constructor is automatically annotated with @JsonCreator
-        }
-    }
-
-    // The path to the model is used as the key in the map because it is unique to
-    // the model, and should not change
-    @JsonProperty("modelPathToProperties")
-    private HashMap<Path, ModelProperties> modelPathToProperties =
-            new HashMap<Path, ModelProperties>();
-
-    /**
-     * Constructor for the NeuralNetworkProperties class.
-     *
-     * <p>This object holds a LinkedList of {@link ModelProperties} objects
-     */
-    public NeuralNetworkPropertyManager() {}
-
-    /**
-     * Constructor for the NeuralNetworkProperties class.
-     *
-     * <p>This object holds a LinkedList of {@link ModelProperties} objects.
-     *
-     * @param modelPropertiesList When the class is constructed, it will hold the provided list
-     */
-    public NeuralNetworkPropertyManager(HashMap<Path, ModelProperties> modelPropertiesList) {}
-
-    @Override
-    public String toString() {
-        String toReturn = "";
-
-        toReturn += "NeuralNetworkProperties [";
-
-        toReturn += modelPathToProperties.toString() + "]";
-
-        return toReturn;
-    }
-
-    /**
-     * Add a model to the list of models.
-     *
-     * @param modelProperties
-     */
-    public void addModelProperties(ModelProperties modelProperties) {
-        modelPathToProperties.put(modelProperties.modelPath, modelProperties);
-    }
-
-    /**
-     * Add two Neural Network Properties together.
-     *
-     * <p>Any properties that are the same will be overwritten by the second
-     *
-     * @param nnProps
-     * @return itself, so it can be chained and used fluently
-     */
-    public NeuralNetworkPropertyManager sum(NeuralNetworkPropertyManager nnProps) {
-        modelPathToProperties.putAll(nnProps.modelPathToProperties);
-
-        return this;
-    }
-
-    /**
-     * Remove a model from the list of models.
-     *
-     * @param modelPath
-     * @return True if the model was removed, false if it was not found
-     */
-    public boolean removeModel(Path modelPath) {
-        return modelPathToProperties.remove(modelPath) != null;
-    }
-
-    /**
-     * Get the model properties for a given model path.
-     *
-     * @param modelPath
-     * @return {@link ModelProperties} object
-     */
-    public ModelProperties getModel(Path modelPath) {
-        return modelPathToProperties.get(modelPath);
-    }
-
-    /**
-     * Get all models
-     *
-     * @return A list of all models
-     */
-    @JsonIgnore
-    public ModelProperties[] getModels() {
-        return modelPathToProperties.values().toArray(new ModelProperties[0]);
-    }
-
-    /**
-     * Change the nickname of a {@link ModelProperties} object.
-     *
-     * @param modelPath
-     * @param newName
-     * @return True if the model was found and renamed, false if it was not found
-     */
-    public boolean renameModel(Path modelPath, String newName) {
-        ModelProperties temp = modelPathToProperties.get(modelPath);
-        if (temp != null) {
-            modelPathToProperties.remove(modelPath);
-            modelPathToProperties.put(
-                    modelPath,
-                    new ModelProperties(
-                            temp.modelPath,
-                            newName,
-                            temp.labels,
-                            temp.resolutionWidth,
-                            temp.resolutionHeight,
-                            temp.family,
-                            temp.version));
-            return true;
-        }
-        return false;
-    }
-
-    public boolean clear() {
-        modelPathToProperties.clear();
-        return true;
-    }
-}

photon-core/src/main/java/org/photonvision/common/configuration/PhotonConfiguration.java

@@ -28,7 +28,7 @@ public class PhotonConfiguration {
     private final HardwareSettings hardwareSettings;
     private NetworkConfig networkConfig;
     private AprilTagFieldLayout atfl;
-    private NeuralNetworkPropertyManager neuralNetworkProperties;
+    private NeuralNetworkModelsSettings neuralNetworkProperties;
     private HashMap<String, CameraConfiguration> cameraConfigurations;
 
     public PhotonConfiguration(
@@ -36,7 +36,7 @@ public class PhotonConfiguration {
             HardwareSettings hardwareSettings,
             NetworkConfig networkConfig,
             AprilTagFieldLayout atfl,
-            NeuralNetworkPropertyManager neuralNetworkProperties) {
+            NeuralNetworkModelsSettings neuralNetworkProperties) {
         this(
                 hardwareConfig,
                 hardwareSettings,
@@ -51,7 +51,7 @@ public class PhotonConfiguration {
             HardwareSettings hardwareSettings,
             NetworkConfig networkConfig,
             AprilTagFieldLayout atfl,
-            NeuralNetworkPropertyManager neuralNetworkProperties,
+            NeuralNetworkModelsSettings neuralNetworkProperties,
             HashMap<String, CameraConfiguration> cameraConfigurations) {
         this.hardwareConfig = hardwareConfig;
         this.hardwareSettings = hardwareSettings;
@@ -67,7 +67,7 @@ public class PhotonConfiguration {
                 new HardwareSettings(),
                 new NetworkConfig(),
                 new AprilTagFieldLayout(List.of(), 0, 0),
-                new NeuralNetworkPropertyManager());
+                new NeuralNetworkModelsSettings());
     }
 
     public HardwareConfig getHardwareConfig() {
@@ -86,7 +86,7 @@ public class PhotonConfiguration {
         return atfl;
     }
 
-    public NeuralNetworkPropertyManager neuralNetworkPropertyManager() {
+    public NeuralNetworkModelsSettings neuralNetworkPropertyManager() {
         return neuralNetworkProperties;
     }
 
@@ -98,7 +98,7 @@ public class PhotonConfiguration {
         this.networkConfig = networkConfig;
     }
 
-    public void setNeuralNetworkProperties(NeuralNetworkPropertyManager neuralNetworkProperties) {
+    public void setNeuralNetworkProperties(NeuralNetworkModelsSettings neuralNetworkProperties) {
         this.neuralNetworkProperties = neuralNetworkProperties;
     }
 
@@ -132,6 +132,16 @@ public class PhotonConfiguration {
 
     @Override
     public String toString() {
+        StringBuilder cameraConfigurationsString = new StringBuilder();
+        cameraConfigurations.forEach(
+                (key, value) -> {
+                    cameraConfigurationsString
+                            .append("\n    ")
+                            .append(key)
+                            .append(" -> ")
+                            .append(value.toString());
+                });
+
         return "PhotonConfiguration [\n  hardwareConfig="
                 + hardwareConfig
                 + "\n  hardwareSettings="
@@ -142,8 +152,8 @@ public class PhotonConfiguration {
                 + atfl
                 + "\n  neuralNetworkProperties="
                 + neuralNetworkProperties
-                + "\n  cameraConfigurations="
-                + cameraConfigurations
-                + "\n]";
+                + "\n  cameraConfigurations={"
+                + cameraConfigurationsString
+                + "}\n]";
     }
 }

photon-core/src/main/java/org/photonvision/common/configuration/SqlConfigProvider.java

@@ -269,7 +269,7 @@ public class SqlConfigProvider extends ConfigProvider {
         } else {
             logger.debug("No " + ref.getSimpleName() + " in database");
         }
-        // either the config entry is empty or Jackson threw and exception
+        // either the config entry is empty or Jackson threw an exception
         try {
             configObj = factory.get();
             logger.info("Loaded default " + ref.getSimpleName());
@@ -313,8 +313,8 @@ public class SqlConfigProvider extends ConfigProvider {
                     loadConfigOrDefault(
                             conn,
                             GlobalKeys.NEURAL_NETWORK_PROPERTIES,
-                            NeuralNetworkPropertyManager.class,
-                            NeuralNetworkPropertyManager::new);
+                            NeuralNetworkModelsSettings.class,
+                            NeuralNetworkModelsSettings::new);
             var atfl =
                     loadConfigOrDefault(
                             conn, GlobalKeys.ATFL_CONFIG_FILE, AprilTagFieldLayout.class, this::atflDefault);
@@ -612,52 +612,65 @@ public class SqlConfigProvider extends ConfigProvider {
 
             // Iterate over every row/"camera" in the table
             while (result.next()) {
-                List<String> dummyList = new ArrayList<>();
+                String uniqueName = "";
+                try {
+                    List<String> dummyList = new ArrayList<>();
 
-                var uniqueName = result.getString(Columns.CAM_UNIQUE_NAME);
+                    uniqueName = result.getString(Columns.CAM_UNIQUE_NAME);
 
-                // A horrifying hack to keep backward compat with otherpaths
-                // We -really- need to delete this -stupid- otherpaths column. I hate it.
-                var configStr = result.getString(Columns.CAM_CONFIG_JSON);
-                CameraConfiguration config = JacksonUtils.deserialize(configStr, CameraConfiguration.class);
+                    // A horrifying hack to keep backward compat with otherpaths
+                    // We -really- need to delete this -stupid- otherpaths column. I hate it.
+                    var configStr = result.getString(Columns.CAM_CONFIG_JSON);
+                    CameraConfiguration config =
+                            JacksonUtils.deserialize(configStr, CameraConfiguration.class);
 
-                if (config.matchedCameraInfo == null) {
-                    logger.info("Legacy CameraConfiguration detected - upgrading");
+                    if (config.matchedCameraInfo == null) {
+                        logger.info("Legacy CameraConfiguration detected - upgrading");
 
-                    // manually create the matchedCameraInfo ourselves. Need to upgrade:
-                    // baseName, path, otherPaths, cameraType, usbvid/pid -> matchedCameraInfo
-                    config.matchedCameraInfo =
-                            JacksonUtils.deserialize(configStr, LegacyCameraConfigStruct.class).matchedCameraInfo;
+                        // manually create the matchedCameraInfo ourselves. Need to upgrade:
+                        // baseName, path, otherPaths, cameraType, usbvid/pid -> matchedCameraInfo
+                        config.matchedCameraInfo =
+                                JacksonUtils.deserialize(configStr, LegacyCameraConfigStruct.class)
+                                        .matchedCameraInfo;
 
-                    // Except that otherPaths used to be its own column. so hack that in here as well
-                    var otherPaths =
+                        // Except that otherPaths used to be its own column. so hack that in here as well
+                        var otherPaths =
+                                JacksonUtils.deserialize(
+                                        result.getString(Columns.CAM_OTHERPATHS_JSON), String[].class);
+                        if (config.matchedCameraInfo instanceof UsbCameraInfo usbInfo) {
+                            usbInfo.otherPaths = otherPaths;
+                        }
+                    }
+
+                    var driverMode =
                             JacksonUtils.deserialize(
-                                    result.getString(Columns.CAM_OTHERPATHS_JSON), String[].class);
-                    if (config.matchedCameraInfo instanceof UsbCameraInfo usbInfo) {
-                        usbInfo.otherPaths = otherPaths;
+                                    result.getString(Columns.CAM_DRIVERMODE_JSON), DriverModePipelineSettings.class);
+                    List<?> pipelineSettings =
+                            JacksonUtils.deserialize(
+                                    result.getString(Columns.CAM_PIPELINE_JSONS), dummyList.getClass());
+
+                    List<CVPipelineSettings> loadedSettings = new ArrayList<>();
+                    for (var setting : pipelineSettings) {
+                        if (setting instanceof String str) {
+                            try {
+                                loadedSettings.add(JacksonUtils.deserialize(str, CVPipelineSettings.class));
+                            } catch (IOException e) {
+                                logger.error(
+                                        "Could not deserialize pipeline setting for camera " + config.nickname, e);
+                            }
+                        }
                     }
+
+                    config.pipelineSettings = loadedSettings;
+                    config.driveModeSettings = driverMode;
+                    loadedConfigurations.put(uniqueName, config);
+                } catch (IOException e) {
+                    logger.error(
+                            "Could not deserialize camera configuration " + uniqueName + " from database!", e);
                 }
-
-                var driverMode =
-                        JacksonUtils.deserialize(
-                                result.getString(Columns.CAM_DRIVERMODE_JSON), DriverModePipelineSettings.class);
-                List<?> pipelineSettings =
-                        JacksonUtils.deserialize(
-                                result.getString(Columns.CAM_PIPELINE_JSONS), dummyList.getClass());
-
-                List<CVPipelineSettings> loadedSettings = new ArrayList<>();
-                for (var setting : pipelineSettings) {
-                    if (setting instanceof String str) {
-                        loadedSettings.add(JacksonUtils.deserialize(str, CVPipelineSettings.class));
-                    }
-                }
-
-                config.pipelineSettings = loadedSettings;
-                config.driveModeSettings = driverMode;
-                loadedConfigurations.put(uniqueName, config);
             }
-        } catch (SQLException | IOException e) {
-            logger.error("Err loading cameras: ", e);
+        } catch (SQLException e) {
+            logger.error("Err querying database to load cameras: ", e);
         } finally {
             try {
                 if (query != null) query.close();

photon-core/src/main/java/org/photonvision/common/dataflow/websocket/UIGeneralSettings.java

@@ -18,14 +18,14 @@
 package org.photonvision.common.dataflow.websocket;
 
 import java.util.List;
-import org.photonvision.common.configuration.NeuralNetworkPropertyManager;
+import org.photonvision.common.configuration.NeuralNetworkModelsSettings;
 
 public class UIGeneralSettings {
     public UIGeneralSettings(
             String version,
             String gpuAcceleration,
             boolean mrCalWorking,
-            NeuralNetworkPropertyManager.ModelProperties[] availableModels,
+            NeuralNetworkModelsSettings.ModelProperties[] availableModels,
             List<String> supportedBackends,
             String hardwareModel,
             String hardwarePlatform,
@@ -45,7 +45,7 @@ public class UIGeneralSettings {
     public String version;
     public String gpuAcceleration;
     public boolean mrCalWorking;
-    public NeuralNetworkPropertyManager.ModelProperties[] availableModels;
+    public NeuralNetworkModelsSettings.ModelProperties[] availableModels;
     public List<String> supportedBackends;
     public String hardwareModel;
     public String hardwarePlatform;

photon-core/src/main/java/org/photonvision/common/hardware/VisionLED.java

@@ -43,6 +43,7 @@ public class VisionLED implements AutoCloseable {
 
     private VisionLEDMode currentLedMode = VisionLEDMode.kDefault;
     private BooleanSupplier pipelineModeSupplier;
+    private boolean currentOutputState = false;
 
     private float mappedBrightness = 0.0f;
 
@@ -85,10 +86,15 @@ public class VisionLED implements AutoCloseable {
     public void setBrightness(int percentage) {
         mappedBrightness =
                 (float) (MathUtils.map(percentage, 0.0, 100.0, brightnessMin, brightnessMax) / 100.0);
-        setInternal(currentLedMode, false);
+        if (currentOutputState) {
+            for (PwmLed led : dimmableVisionLEDs) {
+                led.setValue(mappedBrightness);
+            }
+        }
     }
 
     public void blink(int pulseLengthMillis, int blinkCount) {
+        TimedTaskManager.getInstance().cancelTask(blinkTaskID);
         blinkImpl(pulseLengthMillis, blinkCount);
         int blinkDuration = pulseLengthMillis * blinkCount * 2;
         TimedTaskManager.getInstance()
@@ -96,19 +102,13 @@ public class VisionLED implements AutoCloseable {
     }
 
     private void blinkImpl(int pulseLengthMillis, int blinkCount) {
-        TimedTaskManager.getInstance().cancelTask(blinkTaskID);
         AtomicInteger blinks = new AtomicInteger();
         TimedTaskManager.getInstance()
                 .addTask(
                         blinkTaskID,
                         () -> {
-                            for (LED led : visionLEDs) {
-                                led.toggle();
-                            }
-                            for (PwmLed led : dimmableVisionLEDs) {
-                                led.setValue(mappedBrightness - led.getValue());
-                            }
-                            if (blinks.incrementAndGet() >= blinkCount * 2) {
+                            setStateImpl(!currentOutputState);
+                            if (blinkCount >= 0 && blinks.incrementAndGet() >= blinkCount * 2) {
                                 TimedTaskManager.getInstance().cancelTask(blinkTaskID);
                             }
                         },
@@ -116,12 +116,16 @@ public class VisionLED implements AutoCloseable {
     }
 
     private void setStateImpl(boolean state) {
-        TimedTaskManager.getInstance().cancelTask(blinkTaskID);
+        currentOutputState = state;
         for (LED led : visionLEDs) {
             led.setOn(state);
         }
         for (PwmLed led : dimmableVisionLEDs) {
-            led.setValue(mappedBrightness);
+            if (state) {
+                led.setValue(mappedBrightness);
+            } else {
+                led.off();
+            }
         }
     }
 
@@ -162,6 +166,7 @@ public class VisionLED implements AutoCloseable {
         var lastLedMode = currentLedMode;
 
         if (fromNT || currentLedMode == VisionLEDMode.kDefault) {
+            TimedTaskManager.getInstance().cancelTask(blinkTaskID);
             switch (newLedMode) {
                 case kDefault -> setStateImpl(pipelineModeSupplier.getAsBoolean());
                 case kOff -> setStateImpl(false);

photon-core/src/main/java/org/photonvision/common/util/TestUtils.java

@@ -228,7 +228,8 @@ public class TestUtils {
         kRobots,
         kTag1_640_480,
         kTag1_16h5_1280,
-        kTag_corner_1280;
+        kTag_corner_1280,
+        k36h11_stress_test;
 
         public final Path path;
 
@@ -237,6 +238,7 @@ public class TestUtils {
             var filename = this.toString().substring(1).toLowerCase();
             var extension = ".jpg";
             if (filename.equals("tag1_16h5_1280")) extension = ".png";
+            if (filename.equals("36h11_stress_test")) extension = ".png";
             return Path.of("apriltag", filename + extension);
         }
 

photon-core/src/main/java/org/photonvision/common/util/file/JacksonUtils.java

@@ -18,6 +18,7 @@
 package org.photonvision.common.util.file;
 
 import com.fasterxml.jackson.core.JsonGenerator;
+import com.fasterxml.jackson.core.StreamReadFeature;
 import com.fasterxml.jackson.core.json.JsonReadFeature;
 import com.fasterxml.jackson.databind.DeserializationContext;
 import com.fasterxml.jackson.databind.DeserializationFeature;
@@ -80,6 +81,7 @@ public class JacksonUtils {
         pathModule.addKeyDeserializer(Path.class, new PathKeyDeserializer());
 
         return JsonMapper.builder()
+                .enable(StreamReadFeature.INCLUDE_SOURCE_IN_LOCATION)
                 .configure(JsonReadFeature.ALLOW_JAVA_COMMENTS, true)
                 .configure(DeserializationFeature.FAIL_ON_UNKNOWN_PROPERTIES, false)
                 .activateDefaultTyping(ptv, ObjectMapper.DefaultTyping.JAVA_LANG_OBJECT)

photon-core/src/main/java/org/photonvision/vision/camera/USBCameras/GenericUSBCameraSettables.java

@@ -167,7 +167,28 @@ public class GenericUSBCameraSettables extends VisionSourceSettables {
         }
     }
 
+    @Override
     public void setAutoExposure(boolean cameraAutoExposure) {
+        if ((configuration.cameraQuirks.hasQuirk(CameraQuirk.ArduOV9281Controls)
+                        || configuration.cameraQuirks.hasQuirk(CameraQuirk.ArduOV9782Controls))
+                && !cameraAutoExposure) {
+            // OV9281 and OV9782 on Linux seems to sometimes ignore our exposure requests on first boot if
+            // we're in manual mode. Poking the camera into and out of auto exposure seems to fix it.
+            try {
+                setAutoExposureImpl(false);
+                Thread.sleep(2000);
+                setAutoExposureImpl(true);
+                Thread.sleep(2000);
+                setAutoExposureImpl(false);
+            } catch (InterruptedException e) {
+                logger.error("Thread interrupted while setting OV9281 or OV9782 exposure!", e);
+            }
+        } else {
+            setAutoExposureImpl(cameraAutoExposure);
+        }
+    }
+
+    public void setAutoExposureImpl(boolean cameraAutoExposure) {
         logger.debug("Setting auto exposure to " + cameraAutoExposure);
 
         if (!cameraAutoExposure) {

photon-core/src/main/java/org/photonvision/vision/frame/Frame.java

@@ -17,11 +17,15 @@
 
 package org.photonvision.vision.frame;
 
+import org.photonvision.common.logging.LogGroup;
+import org.photonvision.common.logging.Logger;
 import org.photonvision.common.util.math.MathUtils;
 import org.photonvision.vision.opencv.CVMat;
 import org.photonvision.vision.opencv.Releasable;
 
 public class Frame implements Releasable {
+    private static final Logger logger = new Logger(Frame.class, LogGroup.General);
+
     public final long sequenceID;
     public final long timestampNanos;
 
@@ -45,6 +49,15 @@ public class Frame implements Releasable {
         this.type = type;
         this.timestampNanos = timestampNanos;
         this.frameStaticProperties = frameStaticProperties;
+
+        logger.trace(
+                () ->
+                        "Allocated Frame "
+                                + sequenceID
+                                + "; color image "
+                                + colorImage.matId
+                                + "; processed "
+                                + processedImage.matId);
     }
 
     public Frame(
@@ -73,6 +86,15 @@ public class Frame implements Releasable {
 
     @Override
     public void release() {
+        logger.trace(
+                () ->
+                        "Releasing Frame "
+                                + sequenceID
+                                + "; color image "
+                                + colorImage.matId
+                                + "; processed "
+                                + processedImage.matId);
+
         colorImage.release();
         processedImage.release();
     }

photon-core/src/main/java/org/photonvision/vision/frame/provider/CpuImageProcessor.java

@@ -24,7 +24,6 @@ import org.photonvision.vision.frame.FrameStaticProperties;
 import org.photonvision.vision.frame.FrameThresholdType;
 import org.photonvision.vision.opencv.CVMat;
 import org.photonvision.vision.opencv.ImageRotationMode;
-import org.photonvision.vision.pipe.CVPipe.CVPipeResult;
 import org.photonvision.vision.pipe.impl.GrayscalePipe;
 import org.photonvision.vision.pipe.impl.HSVPipe;
 import org.photonvision.vision.pipe.impl.RotateImagePipe;
@@ -64,26 +63,18 @@ public abstract class CpuImageProcessor extends FrameProvider {
 
     @Override
     public final Frame get() {
-        // TODO Auto-generated method stub
         var input = getInputMat();
 
+        m_rImagePipe.run(input.colorImage.getMat());
+
         CVMat outputMat = null;
-        long sumNanos = 0;
-
-        {
-            CVPipeResult<Void> out = m_rImagePipe.run(input.colorImage.getMat());
-            sumNanos += out.nanosElapsed;
-        }
-
         if (!input.colorImage.getMat().empty()) {
             if (m_processType == FrameThresholdType.HSV) {
                 var hsvResult = m_hsvPipe.run(input.colorImage.getMat());
                 outputMat = new CVMat(hsvResult.output);
-                sumNanos += hsvResult.nanosElapsed;
             } else if (m_processType == FrameThresholdType.GREYSCALE) {
                 var result = m_grayPipe.run(input.colorImage.getMat());
                 outputMat = new CVMat(result.output);
-                sumNanos += result.nanosElapsed;
             } else {
                 outputMat = new CVMat();
             }

photon-core/src/main/java/org/photonvision/vision/objects/Model.java

@@ -18,7 +18,7 @@
 package org.photonvision.vision.objects;
 
 import org.photonvision.common.configuration.NeuralNetworkModelManager.Family;
-import org.photonvision.common.configuration.NeuralNetworkPropertyManager.ModelProperties;
+import org.photonvision.common.configuration.NeuralNetworkModelsSettings.ModelProperties;
 
 public interface Model {
     public ObjectDetector load();

photon-core/src/main/java/org/photonvision/vision/objects/NullModel.java

@@ -20,7 +20,7 @@ package org.photonvision.vision.objects;
 import java.util.List;
 import org.opencv.core.Mat;
 import org.photonvision.common.configuration.NeuralNetworkModelManager.Family;
-import org.photonvision.common.configuration.NeuralNetworkPropertyManager.ModelProperties;
+import org.photonvision.common.configuration.NeuralNetworkModelsSettings.ModelProperties;
 import org.photonvision.vision.pipe.impl.NeuralNetworkPipeResult;
 
 /**

photon-core/src/main/java/org/photonvision/vision/objects/RknnModel.java

@@ -21,7 +21,7 @@ import java.io.File;
 import org.opencv.core.Size;
 import org.photonvision.common.configuration.NeuralNetworkModelManager.Family;
 import org.photonvision.common.configuration.NeuralNetworkModelManager.Version;
-import org.photonvision.common.configuration.NeuralNetworkPropertyManager.ModelProperties;
+import org.photonvision.common.configuration.NeuralNetworkModelsSettings.ModelProperties;
 
 public class RknnModel implements Model {
     public final File modelFile;

photon-core/src/main/java/org/photonvision/vision/objects/RubikModel.java

@@ -21,7 +21,7 @@ import java.io.File;
 import org.opencv.core.Size;
 import org.photonvision.common.configuration.NeuralNetworkModelManager.Family;
 import org.photonvision.common.configuration.NeuralNetworkModelManager.Version;
-import org.photonvision.common.configuration.NeuralNetworkPropertyManager.ModelProperties;
+import org.photonvision.common.configuration.NeuralNetworkModelsSettings.ModelProperties;
 
 public class RubikModel implements Model {
     public final File modelFile;

photon-core/src/main/java/org/photonvision/vision/opencv/CVMat.java

@@ -18,21 +18,57 @@
 package org.photonvision.vision.opencv;
 
 import edu.wpi.first.util.RawFrame;
-import java.util.HashMap;
+import java.lang.ref.PhantomReference;
+import java.lang.ref.ReferenceQueue;
+import java.util.Collections;
+import java.util.Set;
+import java.util.concurrent.ConcurrentHashMap;
+import java.util.concurrent.atomic.AtomicInteger;
 import org.opencv.core.Mat;
 import org.photonvision.common.logging.LogGroup;
 import org.photonvision.common.logging.Logger;
 
 public class CVMat implements Releasable {
     private static final Logger logger = new Logger(CVMat.class, LogGroup.General);
+    private static final AtomicInteger matIdCounter = new AtomicInteger(0);
 
-    private static int allMatCounter = 0;
-    private static final HashMap<Mat, Integer> allMats = new HashMap<>();
+    // All mats that have not yet been released(). these may still need to be GCed
+    private static final Set<MatTracker> allMats =
+            Collections.newSetFromMap(new ConcurrentHashMap<>());
+    private static final ReferenceQueue<CVMat> refQueue = new ReferenceQueue<>();
 
     private static boolean shouldPrint;
 
-    private final Mat mat;
-    private final RawFrame backingFrame;
+    private Mat mat;
+    private RawFrame backingFrame;
+    public final int matId;
+    private final MatTracker tracker;
+    private volatile boolean released = false;
+
+    /** Track a single CVMat instance using a PhantomReference */
+    private static class MatTracker extends PhantomReference<CVMat> {
+        final int id;
+        final long nativePtr;
+        final String allocTrace;
+        volatile boolean explicitlyReleased = false;
+
+        MatTracker(CVMat cvmat, int id, ReferenceQueue<CVMat> queue) {
+            super(cvmat, queue);
+            this.id = id;
+            this.nativePtr = cvmat.mat.nativeObj;
+            this.allocTrace = shouldPrint ? getStackTrace() : "";
+        }
+
+        private static String getStackTrace() {
+            var trace = Thread.currentThread().getStackTrace();
+            final int SKIP = 4; // Skip getStackTrace, <init>, CVMat.<init>, caller
+            var sb = new StringBuilder();
+            for (int i = SKIP; i < Math.min(trace.length, SKIP + 10); i++) {
+                sb.append("\n\t").append(trace[i]);
+            }
+            return sb.toString();
+        }
+    }
 
     public CVMat() {
         this(new Mat());
@@ -42,6 +78,19 @@ public class CVMat implements Releasable {
         this(mat, null);
     }
 
+    public CVMat(Mat mat, RawFrame frame) {
+        this.mat = mat;
+        this.backingFrame = frame;
+        this.matId = matIdCounter.incrementAndGet();
+        this.tracker = new MatTracker(this, matId, refQueue);
+
+        allMats.add(tracker);
+
+        if (shouldPrint) {
+            logger.trace("CVMat" + matId + " allocated - count: " + allMats.size() + tracker.allocTrace);
+        }
+    }
+
     public void copyFrom(CVMat srcMat) {
         copyFrom(srcMat.getMat());
     }
@@ -50,56 +99,73 @@ public class CVMat implements Releasable {
         srcMat.copyTo(mat);
     }
 
-    private StringBuilder getStackTraceBuilder() {
-        var trace = Thread.currentThread().getStackTrace();
-
-        final int STACK_FRAMES_TO_SKIP = 3;
-        final var traceStr = new StringBuilder();
-        for (int idx = STACK_FRAMES_TO_SKIP; idx < trace.length; idx++) {
-            traceStr.append("\t\n").append(trace[idx]);
-        }
-        traceStr.append("\n");
-        return traceStr;
-    }
-
-    public CVMat(Mat mat, RawFrame frame) {
-        this.mat = mat;
-        this.backingFrame = frame;
-
-        allMatCounter++;
-        allMats.put(mat, allMatCounter);
-
-        if (shouldPrint) {
-            logger.trace(() -> "CVMat" + allMatCounter + " alloc - new count: " + allMats.size());
-            logger.trace(getStackTraceBuilder()::toString);
-        }
-    }
-
     @Override
     public void release() {
-        if (this.backingFrame != null) this.backingFrame.close();
+        synchronized (this) {
+            if (released) {
+                if (shouldPrint) {
+                    logger.error("CVMat" + matId + " already released (ignored)");
+                }
+                return;
+            }
+            released = true;
+        }
 
-        // If this mat is empty, all we can do is return
-        if (mat.empty()) return;
+        tracker.explicitlyReleased = true;
 
-        // If the mat isn't in the hashmap, we can't remove it
-        Integer matNo = allMats.get(mat);
-        if (matNo != null) allMats.remove(mat);
-        mat.release();
+        // Free RawFrames exactly ONCE
+        if (backingFrame != null) {
+            try {
+                backingFrame.close();
+                backingFrame = null;
+            } catch (Exception e) {
+                logger.error("Error closing RawFrame for CVMat" + matId, e);
+            }
+        }
+
+        try {
+            if (mat != null) {
+                mat.release();
+                mat = null;
+            } else {
+                logger.error("Mat was already null, this is a no-op");
+            }
+        } catch (Exception e) {
+            logger.error("Error releasing Mat for CVMat" + matId, e);
+        }
+
+        // write down it's freed
+        allMats.remove(tracker);
 
         if (shouldPrint) {
-            logger.trace(() -> "CVMat" + matNo + " de-alloc - new count: " + allMats.size());
-            logger.trace(getStackTraceBuilder()::toString);
+            logger.trace("CVMat" + matId + " released - count: " + allMats.size());
         }
     }
 
     public Mat getMat() {
+        if (released) {
+            throw new IllegalStateException("CVMat" + matId + " has been released!");
+        }
         return mat;
     }
 
+    public boolean isReleased() {
+        return released;
+    }
+
     @Override
     public String toString() {
-        return "CVMat{" + mat.toString() + '}';
+        return "CVMat [mat="
+                + mat
+                + ", backingFrame="
+                + backingFrame
+                + ", matId="
+                + matId
+                + ", tracker="
+                + tracker
+                + ", released="
+                + released
+                + "]";
     }
 
     public static int getMatCount() {
@@ -109,4 +175,61 @@ public class CVMat implements Releasable {
     public static void enablePrint(boolean enabled) {
         shouldPrint = enabled;
     }
+
+    // todo move to somewhere else
+    static {
+        Thread cleanupThread =
+                new Thread(
+                        () -> {
+                            while (true) {
+                                try {
+                                    MatTracker ref = (MatTracker) refQueue.remove();
+
+                                    // Check if it was released before GC
+                                    if (!ref.explicitlyReleased) {
+                                        // This is a leak - remove from tracking and warn
+                                        allMats.remove(ref);
+
+                                        logger.warn(
+                                                "CVMat"
+                                                        + ref.id
+                                                        + " was GC'd without release()! "
+                                                        + "Native memory may have leaked."
+                                                        + "\nAllocated by "
+                                                        + ref.allocTrace);
+                                        if (ref.allocTrace != null) {
+                                            logger.warn("Allocated at:" + ref.allocTrace);
+                                        }
+                                    }
+
+                                    // Because we use PhantomReferences, we can't try to be nice
+                                } catch (InterruptedException e) {
+                                    Thread.currentThread().interrupt();
+                                    break;
+                                }
+                            }
+                        },
+                        "CVMat-Cleanup");
+        cleanupThread.setDaemon(true);
+        cleanupThread.start();
+    }
+
+    // Paranoia
+    @Override
+    @SuppressWarnings("deprecation")
+    protected void finalize() throws Throwable {
+        try {
+            if (!released) {
+                logger.error(
+                        "CVMat"
+                                + matId
+                                + " finalized without release()! Leaking native memory. Allocated by "
+                                + tracker.allocTrace);
+                // Don't call release() here - finalization order is unpredictable
+                // and backingFrame might already be finalized
+            }
+        } finally {
+            super.finalize();
+        }
+    }
 }

photon-core/src/main/java/org/photonvision/vision/pipe/impl/Draw2dAprilTagsPipe.java

@@ -23,8 +23,8 @@ import org.photonvision.vision.frame.FrameDivisor;
 public class Draw2dAprilTagsPipe extends Draw2dTargetsPipe {
     public static class Draw2dAprilTagsParams extends Draw2dTargetsPipe.Draw2dTargetsParams {
         public Draw2dAprilTagsParams(
-                boolean shouldDraw, boolean showMultipleTargets, FrameDivisor divisor) {
-            super(shouldDraw, showMultipleTargets, divisor);
+                boolean shouldDraw, int outputMaximumTargets, FrameDivisor divisor) {
+            super(shouldDraw, outputMaximumTargets, divisor);
             // We want to show the polygon, not the rotated box
             this.showRotatedBox = false;
             this.showMaximumBox = false;

photon-core/src/main/java/org/photonvision/vision/pipe/impl/Draw2dArucoPipe.java

@@ -22,9 +22,8 @@ import org.photonvision.vision.frame.FrameDivisor;
 
 public class Draw2dArucoPipe extends Draw2dTargetsPipe {
     public static class Draw2dArucoParams extends Draw2dTargetsPipe.Draw2dTargetsParams {
-        public Draw2dArucoParams(
-                boolean shouldDraw, boolean showMultipleTargets, FrameDivisor divisor) {
-            super(shouldDraw, showMultipleTargets, divisor);
+        public Draw2dArucoParams(boolean shouldDraw, int outputMaximumTargets, FrameDivisor divisor) {
+            super(shouldDraw, outputMaximumTargets, divisor);
             // We want to show the polygon, not the rotated box
             this.showRotatedBox = false;
             this.showMaximumBox = false;

photon-core/src/main/java/org/photonvision/vision/pipe/impl/Draw2dTargetsPipe.java

@@ -58,14 +58,10 @@ public class Draw2dTargetsPipe
             var circleColor = ColorHelper.colorToScalar(params.circleColor);
             var shapeColour = ColorHelper.colorToScalar(params.shapeOutlineColour);
 
-            for (int i = 0; i < (params.showMultipleTargets ? in.getSecond().size() : 1); i++) {
+            for (int i = 0; i < Math.min(params.outputMaximumTargets, in.getSecond().size()); i++) {
                 Point[] vertices = new Point[4];
                 MatOfPoint contour = new MatOfPoint();
 
-                if (i != 0 && !params.showMultipleTargets) {
-                    break;
-                }
-
                 TrackedTarget target = in.getSecond().get(i);
                 RotatedRect r = target.getMinAreaRect();
 
@@ -233,8 +229,7 @@ public class Draw2dTargetsPipe
         public Color shapeOutlineColour = Color.MAGENTA;
         public Color textColor = Color.GREEN;
         public Color circleColor = Color.RED;
-
-        public final boolean showMultipleTargets;
+        public int outputMaximumTargets;
         public final boolean shouldDraw;
 
         public final FrameDivisor divisor;
@@ -247,10 +242,9 @@ public class Draw2dTargetsPipe
             return shape != null && shape.shape.equals(ContourShape.Circle);
         }
 
-        public Draw2dTargetsParams(
-                boolean shouldDraw, boolean showMultipleTargets, FrameDivisor divisor) {
+        public Draw2dTargetsParams(boolean shouldDraw, int outputMaximumTargets, FrameDivisor divisor) {
             this.shouldDraw = shouldDraw;
-            this.showMultipleTargets = showMultipleTargets;
+            this.outputMaximumTargets = outputMaximumTargets;
             this.divisor = divisor;
         }
     }

photon-core/src/main/java/org/photonvision/vision/pipe/impl/FocusPipe.java

@@ -25,7 +25,9 @@ import org.opencv.imgproc.Imgproc;
 import org.photonvision.vision.pipe.CVPipe;
 
 public class FocusPipe extends CVPipe<Mat, FocusPipe.FocusResult, FocusPipe.FocusParams> {
-    private double maxVariance = 0.0;
+    // cache these
+    MatOfDouble mean = new MatOfDouble();
+    MatOfDouble stddev = new MatOfDouble();
 
     @Override
     protected FocusResult process(Mat in) {
@@ -33,8 +35,6 @@ public class FocusPipe extends CVPipe<Mat, FocusPipe.FocusResult, FocusPipe.Focu
 
         Imgproc.Laplacian(in, outputMat, CvType.CV_64F, 3);
 
-        var mean = new MatOfDouble();
-        var stddev = new MatOfDouble();
         Core.meanStdDev(outputMat, mean, stddev);
         var sd = stddev.get(0, 0)[0];
         var variance = sd * sd;

photon-core/src/main/java/org/photonvision/vision/pipeline/AdvancedPipelineSettings.java

@@ -17,6 +17,7 @@
 
 package org.photonvision.vision.pipeline;
 
+import com.fasterxml.jackson.annotation.JsonAnySetter;
 import java.util.Objects;
 import org.opencv.core.Point;
 import org.photonvision.common.util.numbers.DoubleCouple;
@@ -41,7 +42,7 @@ public class AdvancedPipelineSettings extends CVPipelineSettings {
     public boolean hueInverted = false;
 
     public boolean outputShouldDraw = true;
-    public boolean outputShowMultipleTargets = false;
+    public int outputMaximumTargets = 20;
 
     public DoubleCouple contourArea = new DoubleCouple(0.0, 100.0);
     public DoubleCouple contourRatio = new DoubleCouple(0.0, 20.0);
@@ -90,6 +91,22 @@ public class AdvancedPipelineSettings extends CVPipelineSettings {
     public int cornerDetectionSideCount = 4;
     public double cornerDetectionAccuracyPercentage = 10;
 
+    /**
+     * Handles backward compatibility for the deprecated outputShowMultipleTargets property. When
+     * outputShowMultipleTargets is encountered during deserialization, it sets outputMaximumTargets
+     * appropriately. If outputShowMultipleTargets is false, outputMaximumTargets is set to 1.
+     */
+    @JsonAnySetter
+    public void handleUnknownProperty(String name, Object value) {
+        // Handle the old showMultipleTargets property for backward compatibility
+        if ("outputShowMultipleTargets".equals(name) && value instanceof Boolean showMultipleTargets) {
+            if (!showMultipleTargets) {
+                // If showMultipleTargets is false, limit to 1 target
+                outputMaximumTargets = 1;
+            }
+        }
+    }
+
     @Override
     public boolean equals(Object o) {
         if (this == o) return true;
@@ -97,7 +114,7 @@ public class AdvancedPipelineSettings extends CVPipelineSettings {
         if (!super.equals(o)) return false;
         AdvancedPipelineSettings that = (AdvancedPipelineSettings) o;
         return outputShouldDraw == that.outputShouldDraw
-                && outputShowMultipleTargets == that.outputShowMultipleTargets
+                && outputMaximumTargets == that.outputMaximumTargets
                 && contourSpecklePercentage == that.contourSpecklePercentage
                 && Double.compare(that.offsetDualPointAArea, offsetDualPointAArea) == 0
                 && Double.compare(that.offsetDualPointBArea, offsetDualPointBArea) == 0
@@ -136,7 +153,7 @@ public class AdvancedPipelineSettings extends CVPipelineSettings {
                 hsvValue,
                 hueInverted,
                 outputShouldDraw,
-                outputShowMultipleTargets,
+                outputMaximumTargets,
                 contourArea,
                 contourRatio,
                 contourFullness,

photon-core/src/main/java/org/photonvision/vision/pipeline/AprilTagPipeline.java

@@ -30,6 +30,9 @@ import java.util.ArrayList;
 import java.util.List;
 import java.util.Optional;
 import org.photonvision.common.configuration.ConfigManager;
+import org.photonvision.common.dataflow.structures.Packet;
+import org.photonvision.common.logging.LogGroup;
+import org.photonvision.common.logging.Logger;
 import org.photonvision.common.util.math.MathUtils;
 import org.photonvision.estimation.TargetModel;
 import org.photonvision.targeting.MultiTargetPNPResult;
@@ -49,6 +52,8 @@ import org.photonvision.vision.target.TrackedTarget;
 import org.photonvision.vision.target.TrackedTarget.TargetCalculationParameters;
 
 public class AprilTagPipeline extends CVPipeline<CVPipelineResult, AprilTagPipelineSettings> {
+    private static final Logger logger = new Logger(AprilTagPipeline.class, LogGroup.VisionModule);
+
     private final AprilTagDetectionPipe aprilTagDetectionPipe = new AprilTagDetectionPipe();
     private final AprilTagPoseEstimatorPipe singleTagPoseEstimatorPipe =
             new AprilTagPoseEstimatorPipe();
@@ -232,6 +237,12 @@ public class AprilTagPipeline extends CVPipeline<CVPipelineResult, AprilTagPipel
             }
         }
 
+        if (targetList.size() > Packet.MAX_ARRAY_LEN) {
+            logger.error(
+                    "We have " + targetList.size() + " targets! Arbitrarily dropping some on the floor");
+            targetList = targetList.subList(0, Packet.MAX_ARRAY_LEN);
+        }
+
         var fpsResult = calculateFPSPipe.run(null);
         var fps = fpsResult.output;
 

photon-core/src/main/java/org/photonvision/vision/pipeline/AprilTagPipelineSettings.java

@@ -40,7 +40,6 @@ public class AprilTagPipelineSettings extends AdvancedPipelineSettings {
     public AprilTagPipelineSettings() {
         super();
         pipelineType = PipelineType.AprilTag;
-        outputShowMultipleTargets = true;
         targetModel = TargetModel.kAprilTag6p5in_36h11;
         cameraExposureRaw = 20;
         cameraAutoExposure = false;

photon-core/src/main/java/org/photonvision/vision/pipeline/ArucoPipeline.java

@@ -15,23 +15,6 @@
  * along with this program.  If not, see <https://www.gnu.org/licenses/>.
  */
 
-/*
- * Copyright (C) Photon Vision.
- *
- * This program is free software: you can redistribute it and/or modify
- * it under the terms of the GNU General Public License as published by
- * the Free Software Foundation, either version 3 of the License, or
- * (at your option) any later version.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program.  If not, see <https://www.gnu.org/licenses/>.
- */
-
 package org.photonvision.vision.pipeline;
 
 import edu.wpi.first.apriltag.AprilTagPoseEstimate;
@@ -46,6 +29,9 @@ import org.opencv.core.Mat;
 import org.opencv.imgproc.Imgproc;
 import org.opencv.objdetect.Objdetect;
 import org.photonvision.common.configuration.ConfigManager;
+import org.photonvision.common.dataflow.structures.Packet;
+import org.photonvision.common.logging.LogGroup;
+import org.photonvision.common.logging.Logger;
 import org.photonvision.common.util.math.MathUtils;
 import org.photonvision.estimation.TargetModel;
 import org.photonvision.targeting.MultiTargetPNPResult;
@@ -61,6 +47,8 @@ import org.photonvision.vision.target.TrackedTarget;
 import org.photonvision.vision.target.TrackedTarget.TargetCalculationParameters;
 
 public class ArucoPipeline extends CVPipeline<CVPipelineResult, ArucoPipelineSettings> {
+    private static final Logger logger = new Logger(ArucoPipeline.class, LogGroup.VisionModule);
+
     private ArucoDetectionPipe arucoDetectionPipe = new ArucoDetectionPipe();
     private ArucoPoseEstimatorPipe singleTagPoseEstimatorPipe = new ArucoPoseEstimatorPipe();
     private final MultiTargetPNPPipe multiTagPNPPipe = new MultiTargetPNPPipe();
@@ -237,6 +225,12 @@ public class ArucoPipeline extends CVPipeline<CVPipelineResult, ArucoPipelineSet
             }
         }
 
+        if (targetList.size() > Packet.MAX_ARRAY_LEN) {
+            logger.error(
+                    "We have " + targetList.size() + " targets! Arbitrarily dropping some on the floor");
+            targetList = targetList.subList(0, Packet.MAX_ARRAY_LEN);
+        }
+
         var fpsResult = calculateFPSPipe.run(null);
         var fps = fpsResult.output;
 

photon-core/src/main/java/org/photonvision/vision/pipeline/ArucoPipelineSettings.java

@@ -45,7 +45,6 @@ public class ArucoPipelineSettings extends AdvancedPipelineSettings {
     public ArucoPipelineSettings() {
         super();
         pipelineType = PipelineType.Aruco;
-        outputShowMultipleTargets = true;
         targetModel = TargetModel.kAprilTag6p5in_36h11;
         cameraExposureRaw = 20;
         cameraAutoExposure = true;

photon-core/src/main/java/org/photonvision/vision/pipeline/CVPipeline.java

@@ -26,8 +26,6 @@ import org.photonvision.vision.pipeline.result.CVPipelineResult;
 
 public abstract class CVPipeline<R extends CVPipelineResult, S extends CVPipelineSettings>
         implements Releasable {
-    static final int MAX_MULTI_TARGET_RESULTS = 10;
-
     protected S settings;
     protected FrameStaticProperties frameStaticProperties;
     protected QuirkyCamera cameraQuirks;

photon-core/src/main/java/org/photonvision/vision/pipeline/ColoredShapePipeline.java

@@ -101,11 +101,7 @@ public class ColoredShapePipeline
 
         sortContoursPipe.setParams(
                 new SortContoursPipe.SortContoursParams(
-                        settings.contourSortMode,
-                        settings.outputShowMultipleTargets
-                                ? MAX_MULTI_TARGET_RESULTS // TODO don't hardcode?
-                                : 1,
-                        frameStaticProperties));
+                        settings.contourSortMode, settings.outputMaximumTargets, frameStaticProperties));
 
         collect2dTargetsPipe.setParams(
                 new Collect2dTargetsPipe.Collect2dTargetsParams(
@@ -131,7 +127,7 @@ public class ColoredShapePipeline
         Draw2dTargetsPipe.Draw2dTargetsParams draw2DTargetsParams =
                 new Draw2dTargetsPipe.Draw2dTargetsParams(
                         settings.outputShouldDraw,
-                        settings.outputShowMultipleTargets,
+                        settings.outputMaximumTargets,
                         settings.streamingFrameDivisor);
         draw2DTargetsParams.showShape = true;
         draw2DTargetsParams.showMaximumBox = false;

photon-core/src/main/java/org/photonvision/vision/pipeline/FocusPipeline.java

@@ -73,6 +73,10 @@ public class FocusPipeline extends CVPipeline<FocusPipelineResult, FocusPipeline
 
         var processedCVMat = new CVMat(displayMat);
 
+        // we no longer need the input frame's processed image, and nobody else will release it if we
+        // don't
+        frame.processedImage.release();
+
         return new FocusPipelineResult(
                 frame.sequenceID,
                 MathUtils.nanosToMillis(totalNanos),

photon-core/src/main/java/org/photonvision/vision/pipeline/ObjectDetectionPipeline.java

@@ -82,9 +82,7 @@ public class ObjectDetectionPipeline
 
         sortContoursPipe.setParams(
                 new SortContoursPipe.SortContoursParams(
-                        settings.contourSortMode,
-                        settings.outputShowMultipleTargets ? MAX_MULTI_TARGET_RESULTS : 1,
-                        frameStaticProperties));
+                        settings.contourSortMode, settings.outputMaximumTargets, frameStaticProperties));
 
         filterContoursPipe.setParams(
                 new FilterObjectDetectionsPipe.FilterContoursParams(

photon-core/src/main/java/org/photonvision/vision/pipeline/ObjectDetectionPipelineSettings.java

@@ -18,18 +18,18 @@
 package org.photonvision.vision.pipeline;
 
 import org.photonvision.common.configuration.NeuralNetworkModelManager;
-import org.photonvision.common.configuration.NeuralNetworkPropertyManager;
+import org.photonvision.common.configuration.NeuralNetworkModelsSettings;
 import org.photonvision.vision.objects.Model;
 
 public class ObjectDetectionPipelineSettings extends AdvancedPipelineSettings {
     public double confidence;
     public double nms; // non maximal suppression
-    public NeuralNetworkPropertyManager.ModelProperties model;
+    public NeuralNetworkModelsSettings.ModelProperties model;
 
     public ObjectDetectionPipelineSettings() {
         super();
         this.pipelineType = PipelineType.ObjectDetection; // TODO: FIX this
-        this.outputShowMultipleTargets = true;
+        this.outputMaximumTargets = 20;
         cameraExposureRaw = 20;
         cameraAutoExposure = false;
         ledMode = false;

photon-core/src/main/java/org/photonvision/vision/pipeline/OutputStreamPipeline.java

@@ -58,19 +58,19 @@ public class OutputStreamPipeline {
         draw2dTargetsPipe.setParams(
                 new Draw2dTargetsPipe.Draw2dTargetsParams(
                         settings.outputShouldDraw,
-                        settings.outputShowMultipleTargets,
+                        settings.outputMaximumTargets,
                         settings.streamingFrameDivisor));
 
         draw2dAprilTagsPipe.setParams(
                 new Draw2dAprilTagsPipe.Draw2dAprilTagsParams(
                         settings.outputShouldDraw,
-                        settings.outputShowMultipleTargets,
+                        settings.outputMaximumTargets,
                         settings.streamingFrameDivisor));
 
         draw2dArucoPipe.setParams(
                 new Draw2dArucoPipe.Draw2dArucoParams(
                         settings.outputShouldDraw,
-                        settings.outputShowMultipleTargets,
+                        settings.outputMaximumTargets,
                         settings.streamingFrameDivisor));
 
         draw2dCrosshairPipe.setParams(

photon-core/src/main/java/org/photonvision/vision/pipeline/ReflectivePipeline.java

@@ -85,9 +85,7 @@ public class ReflectivePipeline extends CVPipeline<CVPipelineResult, ReflectiveP
 
         sortContoursPipe.setParams(
                 new SortContoursPipe.SortContoursParams(
-                        settings.contourSortMode,
-                        settings.outputShowMultipleTargets ? MAX_MULTI_TARGET_RESULTS : 1,
-                        frameStaticProperties));
+                        settings.contourSortMode, settings.outputMaximumTargets, frameStaticProperties));
 
         collect2dTargetsPipe.setParams(
                 new Collect2dTargetsPipe.Collect2dTargetsParams(

photon-core/src/main/java/org/photonvision/vision/processes/VisionModule.java

@@ -465,13 +465,13 @@ public class VisionModule {
                             pipelineSettings.cameraExposureRaw = 10; // reasonable default
                     }
 
-                    settables.setExposureRaw(pipelineSettings.cameraExposureRaw);
                     try {
                         settables.setAutoExposure(pipelineSettings.cameraAutoExposure);
                     } catch (VideoException e) {
                         logger.error("Unable to set camera auto exposure!");
                         logger.error(e.toString());
                     }
+                    settables.setExposureRaw(pipelineSettings.cameraExposureRaw);
                     if (cameraQuirks.hasQuirk(CameraQuirk.Gain)) {
                         // If the gain is disabled for some reason, re-enable it
                         if (pipelineSettings.cameraGain == -1) pipelineSettings.cameraGain = 75;

photon-core/src/main/java/org/photonvision/vision/processes/VisionModuleChangeSubscriber.java

@@ -25,7 +25,7 @@ import java.util.List;
 import java.util.Map;
 import java.util.concurrent.locks.ReentrantLock;
 import org.opencv.core.Point;
-import org.photonvision.common.configuration.NeuralNetworkPropertyManager.ModelProperties;
+import org.photonvision.common.configuration.NeuralNetworkModelsSettings.ModelProperties;
 import org.photonvision.common.dataflow.DataChangeSubscriber;
 import org.photonvision.common.dataflow.events.DataChangeEvent;
 import org.photonvision.common.dataflow.events.IncomingWebSocketEvent;

photon-core/src/main/java/org/photonvision/vision/processes/VisionRunner.java

@@ -191,8 +191,8 @@ public class VisionRunner {
                 // give up without increasing loop count
                 // Still feed with blank frames just dont run any pipelines
 
+                frame.release();
                 pipelineResultConsumer.accept(new CVPipelineResult(0l, 0, 0, null, new Frame()));
-
             } else if (pipeline == pipelineSupplier.get()) {
                 // If the pipeline has changed while we are getting our frame we should scrap
                 // that frame it may result in incorrect frame settings like hsv values

photon-core/src/test/java/org/photonvision/common/BenchmarkTest.java

@@ -51,7 +51,7 @@ public class BenchmarkTest {
         pipeline.getSettings().hsvSaturation.set(100, 255);
         pipeline.getSettings().hsvValue.set(190, 255);
         pipeline.getSettings().outputShouldDraw = true;
-        pipeline.getSettings().outputShowMultipleTargets = true;
+        pipeline.getSettings().outputMaximumTargets = 20;
         pipeline.getSettings().contourGroupingMode = ContourGroupingMode.Dual;
         pipeline.getSettings().contourIntersection = ContourIntersectionDirection.Up;
 
@@ -105,7 +105,7 @@ public class BenchmarkTest {
         pipeline.getSettings().hsvSaturation.set(100, 255);
         pipeline.getSettings().hsvValue.set(190, 255);
         pipeline.getSettings().outputShouldDraw = true;
-        pipeline.getSettings().outputShowMultipleTargets = true;
+        pipeline.getSettings().outputMaximumTargets = 20;
         pipeline.getSettings().contourGroupingMode = ContourGroupingMode.Dual;
         pipeline.getSettings().contourIntersection = ContourIntersectionDirection.Up;
 

photon-core/src/test/java/org/photonvision/common/ShapeBenchmarkTest.java

@@ -66,7 +66,7 @@ public class ShapeBenchmarkTest {
         pipeline.getSettings().hsvSaturation.set(100, 255);
         pipeline.getSettings().hsvValue.set(190, 255);
         pipeline.getSettings().outputShouldDraw = true;
-        pipeline.getSettings().outputShowMultipleTargets = true;
+        pipeline.getSettings().outputMaximumTargets = 20;
         pipeline.getSettings().contourGroupingMode = ContourGroupingMode.Single;
         pipeline.getSettings().contourIntersection = ContourIntersectionDirection.Up;
         pipeline.getSettings().contourShape = ContourShape.Custom;
@@ -87,7 +87,7 @@ public class ShapeBenchmarkTest {
         pipeline.getSettings().hsvSaturation.set(100, 255);
         pipeline.getSettings().hsvValue.set(190, 255);
         pipeline.getSettings().outputShouldDraw = true;
-        pipeline.getSettings().outputShowMultipleTargets = true;
+        pipeline.getSettings().outputMaximumTargets = 20;
         pipeline.getSettings().contourGroupingMode = ContourGroupingMode.Single;
         pipeline.getSettings().contourIntersection = ContourIntersectionDirection.Up;
         pipeline.getSettings().contourShape = ContourShape.Custom;
@@ -109,7 +109,7 @@ public class ShapeBenchmarkTest {
         pipeline.getSettings().hsvSaturation.set(100, 255);
         pipeline.getSettings().hsvValue.set(190, 255);
         pipeline.getSettings().outputShouldDraw = true;
-        pipeline.getSettings().outputShowMultipleTargets = true;
+        pipeline.getSettings().outputMaximumTargets = 20;
         pipeline.getSettings().contourGroupingMode = ContourGroupingMode.Single;
         pipeline.getSettings().contourIntersection = ContourIntersectionDirection.Up;
         pipeline.getSettings().contourShape = ContourShape.Custom;
@@ -131,7 +131,7 @@ public class ShapeBenchmarkTest {
         pipeline.getSettings().hsvSaturation.set(100, 255);
         pipeline.getSettings().hsvValue.set(190, 255);
         pipeline.getSettings().outputShouldDraw = true;
-        pipeline.getSettings().outputShowMultipleTargets = true;
+        pipeline.getSettings().outputMaximumTargets = 20;
         pipeline.getSettings().contourGroupingMode = ContourGroupingMode.Single;
         pipeline.getSettings().contourIntersection = ContourIntersectionDirection.Up;
         pipeline.getSettings().contourShape = ContourShape.Custom;

photon-core/src/test/java/org/photonvision/common/configuration/NeuralNetworkPropertyManagerTest.java

@@ -25,13 +25,13 @@ import java.util.LinkedList;
 import org.junit.jupiter.api.Test;
 import org.photonvision.common.configuration.NeuralNetworkModelManager.Family;
 import org.photonvision.common.configuration.NeuralNetworkModelManager.Version;
-import org.photonvision.common.configuration.NeuralNetworkPropertyManager.ModelProperties;
+import org.photonvision.common.configuration.NeuralNetworkModelsSettings.ModelProperties;
 import org.photonvision.common.util.file.JacksonUtils;
 
 public class NeuralNetworkPropertyManagerTest {
     @Test
     void testSerialization() {
-        var nnpm = new NeuralNetworkPropertyManager();
+        var nnpm = new NeuralNetworkModelsSettings();
         // Path is always serialized as absolute; for the test to pass, this must also be made absolute
         nnpm.addModelProperties(
                 new ModelProperties(
@@ -45,7 +45,7 @@ public class NeuralNetworkPropertyManagerTest {
         String result = assertDoesNotThrow(() -> JacksonUtils.serializeToString(nnpm));
         var deserializedNnpm =
                 assertDoesNotThrow(
-                        () -> JacksonUtils.deserialize(result, NeuralNetworkPropertyManager.class));
+                        () -> JacksonUtils.deserialize(result, NeuralNetworkModelsSettings.class));
         assertEquals(nnpm.getModels()[0], deserializedNnpm.getModels()[0]);
     }
 }

photon-core/src/test/java/org/photonvision/common/configuration/SQLConfigTest.java

@@ -20,23 +20,32 @@ package org.photonvision.common.configuration;
 import static org.junit.jupiter.api.Assertions.assertDoesNotThrow;
 import static org.junit.jupiter.api.Assertions.assertEquals;
 
+import com.fasterxml.jackson.core.JsonProcessingException;
 import edu.wpi.first.cscore.UsbCameraInfo;
+import java.io.IOException;
 import java.nio.file.Path;
+import java.util.Collection;
 import java.util.List;
+import org.apache.commons.io.FileUtils;
 import org.junit.jupiter.api.BeforeAll;
 import org.junit.jupiter.api.Order;
 import org.junit.jupiter.api.Test;
 import org.junit.jupiter.api.io.TempDir;
 import org.photonvision.common.LoadJNI;
+import org.photonvision.common.configuration.NeuralNetworkModelManager.Family;
 import org.photonvision.common.util.TestUtils;
 import org.photonvision.vision.camera.CameraQuirk;
 import org.photonvision.vision.camera.PVCameraInfo;
+import org.photonvision.vision.pipeline.AdvancedPipelineSettings;
 import org.photonvision.vision.pipeline.AprilTagPipelineSettings;
+import org.photonvision.vision.pipeline.CVPipelineSettings;
 import org.photonvision.vision.pipeline.ColoredShapePipelineSettings;
+import org.photonvision.vision.pipeline.ObjectDetectionPipelineSettings;
+import org.photonvision.vision.pipeline.PipelineType;
 import org.photonvision.vision.pipeline.ReflectivePipelineSettings;
 
 public class SQLConfigTest {
-    @TempDir private static Path tmpDir;
+    @TempDir private Path tmpDir;
 
     @BeforeAll
     public static void init() {
@@ -84,11 +93,15 @@ public class SQLConfigTest {
     }
 
     @Test
-    public void testLoad2024_3_1() {
-        var cfgLoader =
-                new SqlConfigProvider(
-                        TestUtils.getConfigDirectoriesPath(false)
-                                .resolve("photonvision_config_from_v2024.3.1"));
+    public void testLoad2024_3_1() throws IOException {
+        // Copy the 2024.3.1 config to a temp dir
+        FileUtils.copyDirectory(
+                TestUtils.getConfigDirectoriesPath(false)
+                        .resolve("photonvision_config_from_v2024.3.1")
+                        .toFile(),
+                tmpDir.resolve("photonvision_config_from_v2024.3.1").toFile());
+
+        var cfgLoader = new SqlConfigProvider(tmpDir.resolve("photonvision_config_from_v2024.3.1"));
 
         assertDoesNotThrow(cfgLoader::load);
 
@@ -104,4 +117,97 @@ public class SQLConfigTest {
                                     .hasQuirk(c));
         }
     }
+
+    void common2025p3p1Assertions(PhotonConfiguration config) {
+        // Make sure we got 8 cameras
+        assertEquals(8, config.getCameraConfigurations().size());
+
+        // Make sure exactly 2 have object detection pipelines
+        long count =
+                config.getCameraConfigurations().values().stream()
+                        .filter(
+                                c ->
+                                        c.pipelineSettings.stream()
+                                                .anyMatch(s -> s instanceof ObjectDetectionPipelineSettings))
+                        .count();
+        assertEquals(2, count);
+    }
+
+    @Test
+    public void testLoadNewNNMM() throws JsonProcessingException, IOException {
+        var folder = tmpDir.resolve("2025.3.1-old-nnmm");
+        FileUtils.copyDirectory(
+                TestUtils.getConfigDirectoriesPath(false).resolve("2025.3.1-old-nnmm").toFile(),
+                folder.toFile());
+
+        var cfgManager = new ConfigManager(folder, new SqlConfigProvider(folder));
+
+        // Replace global configmanager
+        ConfigManager.INSTANCE = cfgManager;
+
+        assertDoesNotThrow(cfgManager::load);
+
+        System.out.println(cfgManager.getConfig());
+        common2025p3p1Assertions(cfgManager.getConfig());
+
+        // And we now see two models
+        NeuralNetworkModelManager.getInstance();
+        // force us to allow RKNN
+        NeuralNetworkModelManager.getInstance().supportedBackends.add(Family.RKNN);
+        NeuralNetworkModelManager.getInstance().discoverModels();
+        assertEquals(5, NeuralNetworkModelManager.getInstance().models.get(Family.RKNN).size());
+
+        ConfigManager.getInstance().saveToDisk();
+
+        // Now that we have the config saved, load it again
+        var reloadedProvider = new SqlConfigProvider(folder);
+        reloadedProvider.load();
+        common2025p3p1Assertions(reloadedProvider.getConfig());
+
+        // And make sure NNPM has all 5 models
+        assertEquals(5, reloadedProvider.getConfig().neuralNetworkPropertyManager().getModels().length);
+
+        ConfigManager.INSTANCE = null;
+    }
+
+    @Test
+    public void testMaxDetectionsMigration() throws IOException {
+        var folder = tmpDir.resolve("2025.3.1-old-nnmm");
+        FileUtils.copyDirectory(
+                TestUtils.getConfigDirectoriesPath(false).resolve("2025.3.1-old-nnmm").toFile(),
+                folder.toFile());
+
+        var cfgManager = new ConfigManager(folder, new SqlConfigProvider(folder));
+
+        // Replace global configmanager
+        ConfigManager.INSTANCE = cfgManager;
+
+        assertDoesNotThrow(cfgManager::load);
+
+        Collection<CameraConfiguration> cameraConfigs =
+                cfgManager.getConfig().getCameraConfigurations().values();
+
+        for (CameraConfiguration cc : cameraConfigs) {
+            for (CVPipelineSettings ps : cc.pipelineSettings) {
+                if (ps instanceof AdvancedPipelineSettings adps) {
+                    AdvancedPipelineSettings finalPs = adps;
+                    if (finalPs.pipelineType.equals(PipelineType.AprilTag)
+                            || finalPs.pipelineType.equals(PipelineType.Aruco)) {
+                        // Tag pipelines don't have max detections, so skip
+                        continue;
+                    } else if (finalPs.pipelineNickname.equals("TEST MIGRATION")) {
+                        // This is our colored shape pipeline that we set to 1 before saving
+                        assertEquals(1, finalPs.outputMaximumTargets);
+                    } else {
+                        // All others should be at default 20
+                        assertEquals(20, finalPs.outputMaximumTargets);
+                    }
+                } else {
+                    System.out.println("Skipping pipeline settings type: " + ps.getClass().getSimpleName());
+                }
+            }
+        }
+
+        ConfigManager.INSTANCE = null;
+    }
 }

photon-core/src/test/java/org/photonvision/vision/pipeline/AprilTagTest.java

@@ -137,4 +137,31 @@ public class AprilTagTest {
         assertEquals(2, pose.getTranslation().getY(), 0.2);
         assertEquals(0.0, pose.getTranslation().getZ(), 0.2);
     }
+
+    @Test
+    public void testManyDetections() {
+        // Given a 36h11 pipeline
+        var pipeline = new AprilTagPipeline();
+        pipeline.getSettings().inputShouldShow = true;
+        pipeline.getSettings().outputShouldDraw = true;
+        pipeline.getSettings().solvePNPEnabled = true;
+        pipeline.getSettings().cornerDetectionAccuracyPercentage = 4;
+        pipeline.getSettings().cornerDetectionUseConvexHulls = true;
+        pipeline.getSettings().tagFamily = AprilTagFamily.kTag36h11;
+        pipeline.getSettings().outputMaximumTargets = 3; // bogus
+
+        // when we have a picture with 280 targets
+        var frameProvider =
+                new FileFrameProvider(
+                        TestUtils.getApriltagImagePath(TestUtils.ApriltagTestImages.k36h11_stress_test, false),
+                        TestUtils.WPI2020Image.FOV,
+                        TestUtils.getCoeffs(TestUtils.LIMELIGHT_480P_CAL_FILE, false));
+        frameProvider.requestFrameThresholdType(pipeline.getThresholdType());
+
+        CVPipelineResult pipelineResult;
+        pipelineResult = pipeline.run(frameProvider.get(), QuirkyCamera.DefaultCamera);
+
+        // the pipeline will only give us Byte.MAX_VALUE many
+        assertEquals(Byte.MAX_VALUE, pipelineResult.targets.size());
+    }
 }

photon-core/src/test/java/org/photonvision/vision/pipeline/CirclePNPTest.java

@@ -94,7 +94,7 @@ public class CirclePNPTest {
         pipeline.getSettings().cameraCalibration = getCoeffs(LIFECAM_480P_CAL_FILE);
         pipeline.getSettings().targetModel = TargetModel.kCircularPowerCell7in;
         pipeline.getSettings().outputShouldDraw = true;
-        pipeline.getSettings().outputShowMultipleTargets = false;
+        pipeline.getSettings().outputMaximumTargets = 20;
         pipeline.getSettings().contourGroupingMode = ContourGroupingMode.Single;
         pipeline.getSettings().contourIntersection = ContourIntersectionDirection.Up;
         pipeline.getSettings().contourShape = ContourShape.Circle;
@@ -144,7 +144,7 @@ public class CirclePNPTest {
         settings.hsvSaturation.set(100, 255);
         settings.hsvValue.set(190, 255);
         settings.outputShouldDraw = true;
-        settings.outputShowMultipleTargets = true;
+        settings.outputMaximumTargets = 20;
         settings.contourGroupingMode = ContourGroupingMode.Dual;
         settings.contourIntersection = ContourIntersectionDirection.Up;
 

photon-core/src/test/java/org/photonvision/vision/pipeline/ColoredShapePipelineTest.java

@@ -102,7 +102,7 @@ public class ColoredShapePipelineTest {
         settings.hsvSaturation.set(100, 255);
         settings.hsvValue.set(100, 255);
         settings.outputShouldDraw = true;
-        settings.outputShowMultipleTargets = true;
+        settings.outputMaximumTargets = 20;
         settings.contourGroupingMode = ContourGroupingMode.Single;
         settings.contourIntersection = ContourIntersectionDirection.Up;
         settings.contourShape = ContourShape.Triangle;

photon-core/src/test/java/org/photonvision/vision/pipeline/MaxDetectionsTest.java

@@ -0,0 +1,78 @@
+/*
+ * Copyright (C) Photon Vision.
+ *
+ * This program is free software: you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation, either version 3 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <https://www.gnu.org/licenses/>.
+ */
+
+package org.photonvision.vision.pipeline;
+
+import static org.junit.jupiter.api.Assertions.assertEquals;
+
+import java.nio.file.Path;
+import org.junit.jupiter.api.Test;
+import org.photonvision.common.LoadJNI;
+import org.photonvision.common.configuration.ConfigManager;
+import org.photonvision.common.util.TestUtils;
+import org.photonvision.vision.camera.QuirkyCamera;
+import org.photonvision.vision.frame.provider.FileFrameProvider;
+import org.photonvision.vision.opencv.ContourShape;
+import org.photonvision.vision.pipe.impl.HSVPipe;
+import org.photonvision.vision.pipeline.result.CVPipelineResult;
+
+public class MaxDetectionsTest {
+    @Test
+    public void testMaxDetections() {
+        LoadJNI.loadLibraries();
+        ConfigManager.getInstance().load();
+
+        ColoredShapePipeline pipeline = new ColoredShapePipeline();
+
+        pipeline.settings.contourShape = ContourShape.Circle;
+        pipeline.settings.hsvHue.set(140, 160);
+        pipeline.settings.hsvSaturation.set(226, 246);
+        pipeline.settings.hsvValue.set(188, 208);
+        pipeline.settings.maxCannyThresh = 90;
+        pipeline.settings.circleAccuracy = 20;
+        pipeline.settings.circleDetectThreshold = 5;
+
+        Path path =
+                TestUtils.getResourcesFolderPath(false).resolve("testimages/polygons/ColoredShapeTest.png");
+
+        var frameProvider = new FileFrameProvider(path, TestUtils.WPI2019Image.FOV);
+
+        // VisionRunner normally does this
+        var hsvParams =
+                new HSVPipe.HSVParams(
+                        pipeline.getSettings().hsvHue,
+                        pipeline.getSettings().hsvSaturation,
+                        pipeline.getSettings().hsvValue,
+                        pipeline.getSettings().hueInverted);
+        frameProvider.requestHsvSettings(hsvParams);
+        frameProvider.requestFrameThresholdType(pipeline.getThresholdType());
+
+        CVPipelineResult result = pipeline.run(frameProvider.get(), QuirkyCamera.DefaultCamera);
+        TestUtils.showImage(result.inputAndOutputFrame.processedImage.getMat(), "Max Detections Test");
+
+        assertEquals(20, result.targets.size());
+
+        pipeline.settings.outputMaximumTargets = 5;
+        result = pipeline.run(frameProvider.get(), QuirkyCamera.DefaultCamera);
+        assertEquals(5, result.targets.size());
+
+        pipeline.settings.outputMaximumTargets = 50;
+        result = pipeline.run(frameProvider.get(), QuirkyCamera.DefaultCamera);
+        // 24 circles, but we only detect 22
+        assertEquals(22, result.targets.size());
+    }
+}

photon-core/src/test/java/org/photonvision/vision/pipeline/ReflectivePipelineTest.java

@@ -41,7 +41,7 @@ public class ReflectivePipelineTest {
         pipeline.getSettings().hsvSaturation.set(100, 255);
         pipeline.getSettings().hsvValue.set(190, 255);
         pipeline.getSettings().outputShouldDraw = true;
-        pipeline.getSettings().outputShowMultipleTargets = true;
+        pipeline.getSettings().outputMaximumTargets = 20;
         pipeline.getSettings().contourGroupingMode = ContourGroupingMode.Dual;
         pipeline.getSettings().contourIntersection = ContourIntersectionDirection.Up;
 
@@ -120,7 +120,7 @@ public class ReflectivePipelineTest {
         settings.hsvSaturation.set(100, 255);
         settings.hsvValue.set(190, 255);
         settings.outputShouldDraw = true;
-        settings.outputShowMultipleTargets = true;
+        settings.outputMaximumTargets = 20;
         settings.contourGroupingMode = ContourGroupingMode.Dual;
         settings.contourIntersection = ContourIntersectionDirection.Up;
 

photon-core/src/test/java/org/photonvision/vision/pipeline/SolvePNPTest.java

@@ -89,7 +89,6 @@ public class SolvePNPTest {
         pipeline.getSettings().hsvSaturation.set(100, 255);
         pipeline.getSettings().hsvValue.set(190, 255);
         pipeline.getSettings().outputShouldDraw = true;
-        pipeline.getSettings().outputShowMultipleTargets = true;
         pipeline.getSettings().solvePNPEnabled = true;
         pipeline.getSettings().contourGroupingMode = ContourGroupingMode.Dual;
         pipeline.getSettings().contourIntersection = ContourIntersectionDirection.Up;
@@ -225,7 +224,6 @@ public class SolvePNPTest {
         settings.hsvSaturation.set(100, 255);
         settings.hsvValue.set(190, 255);
         settings.outputShouldDraw = true;
-        settings.outputShowMultipleTargets = true;
         settings.contourGroupingMode = ContourGroupingMode.Dual;
         settings.contourIntersection = ContourIntersectionDirection.Up;
 

photon-lib/py/photonlibpy/simulation/photonCameraSim.py

@@ -271,6 +271,9 @@ class PhotonCameraSim:
         camRt = RotTrlTransform3d.makeRelativeTo(cameraPose)
 
         for tgt in targets:
+            if len(detectableTgts) >= 50:
+                break
+
             # pose isn't visible, skip to next
             if not self.canSeeTargetPose(cameraPose, tgt):
                 continue

photon-lib/py/setup.py

@@ -4,15 +4,13 @@ import subprocess
 from setuptools import find_packages, setup
 
 gitDescribeResult = (
-    subprocess.check_output(
-        ["git", "describe", "--tags", "--match=v*", "--exclude=*rc*", "--always"]
-    )
+    subprocess.check_output(["git", "describe", "--tags", "--match=v*", "--always"])
     .decode("utf-8")
     .strip()
 )
 
 m = re.search(
-    r"(v[0-9]{4}\.[0-9]{1}\.[0-9]{1})-?((?:beta)?(?:alpha)?)-?([0-9\.]*)",
+    r"v([0-9]{4}\.[0-9]{1}\.[0-9]{1})-?((?:beta|alpha|rc)?)-?([0-9\.]*)",
     gitDescribeResult,
 )
 
@@ -26,7 +24,7 @@ if m:
         prefix = m.group(1)
         maturity = m.group(2)
         suffix = m.group(3).replace(".", "")
-        versionString = f"{prefix}.{maturity}.{suffix}"
+        versionString = f"{prefix}{maturity}{suffix}"
     else:
         split = gitDescribeResult.split("-")
         if len(split) == 3:
@@ -35,8 +33,7 @@ if m:
             versionString = f"{year[1:]}post{commits}"
             print("using dev release " + versionString)
         else:
-            year = gitDescribeResult
-            versionString = year[1:]
+            versionString = gitDescribeResult[1:]
             print("using full release " + versionString)
 
 
@@ -60,12 +57,12 @@ setup(
     package_data={"photonlibpy": ["py.typed"]},
     version=versionString,
     install_requires=[
-        "numpy~=2.4",
-        "wpilib==2026.1.1",
-        "robotpy-wpimath==2026.1.1",
-        "robotpy-apriltag==2026.1.1",
-        "robotpy-cscore==2026.1.1",
-        "pyntcore==2026.1.1",
+        "numpy~=2.3",
+        "wpilib>=2026.2.1,<2027",
+        "robotpy-wpimath>=2026.2.1,<2027",
+        "robotpy-apriltag>=2026.2.1,<2027",
+        "robotpy-cscore>=2026.2.1,<2027",
+        "pyntcore>=2026.2.1,<2027",
         "opencv-python;platform_machine!='roborio'",
     ],
     description=descriptionStr,

photon-lib/py/test/visionSystemSim_test.py

@@ -78,6 +78,35 @@ def test_VisibilityCupidShuffle() -> None:
     assert camera.getLatestResult().hasTargets()
 
 
+def test_bunchaTargets() -> None:
+    visionSysSim = VisionSystemSim("Test")
+    camera = PhotonCamera("camera")
+    cameraSim = PhotonCameraSim(camera)
+    visionSysSim.addCamera(cameraSim, Transform3d())
+
+    cameraSim.prop.setCalibrationFromFOV(640, 480, fovDiag=Rotation2d.fromDegrees(80.0))
+
+    for i in range(100):
+        targetPose = Pose3d(
+            Translation3d(15.98 + i * 0.1, 0.0, 2.0), Rotation3d(0, 0, math.pi)
+        )
+        visionSysSim.addVisionTargets(
+            [
+                VisionTargetSim(
+                    targetPose,
+                    TargetModel.createPlanar(width=1.0, height=1.0),
+                    4774 + i,
+                )
+            ]
+        )
+
+    robotPose = Pose2d(Translation2d(2.0, 0.0), Rotation2d.fromDegrees(5.0))
+
+    visionSysSim.update(robotPose)
+
+    assert len(camera.getLatestResult().getTargets()) == 50
+
+
 def test_NotVisibleVert1() -> None:
     targetPose = Pose3d(Translation3d(15.98, 0.0, 2.0), Rotation3d(0, 0, math.pi))
 

photon-lib/src/main/java/org/photonvision/PhotonCamera.java

@@ -45,7 +45,6 @@ import edu.wpi.first.wpilibj.Alert;
 import edu.wpi.first.wpilibj.Alert.AlertType;
 import edu.wpi.first.wpilibj.DriverStation;
 import edu.wpi.first.wpilibj.Timer;
-import edu.wpi.first.wpilibj.util.WPILibVersion;
 import java.util.ArrayList;
 import java.util.Arrays;
 import java.util.List;
@@ -189,50 +188,6 @@ public class PhotonCamera implements AutoCloseable {
 
     static void verifyDependencies() {
         // spotless:off
-        if (!WPILibVersion.Version.equals(PhotonVersion.wpilibTargetVersion)) {
-            String bfw = """
-
-
-
-
-                    >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\s
-                    >>> !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\s
-                    >>>                                          \s
-                    >>> You are running an incompatible version  \s
-                    >>> of PhotonVision !                        \s
-                    >>>                                          \s
-                    >>> PhotonLib """
-                    + PhotonVersion.versionString
-                    + " is built for WPILib "
-                    + PhotonVersion.wpilibTargetVersion
-                    + "\n"
-                    + ">>> but you are using WPILib "
-                    + WPILibVersion.Version
-                    + """
-                    \n>>>                                          \s
-                    >>> This is neither tested nor supported.    \s
-                    >>> You MUST update WPILib, PhotonLib, or both.
-                    >>> Check `./gradlew dependencies` and ensure\s
-                    >>> all mentions of OpenCV match the version \s
-                    >>> that PhotonLib was built for. If you find a
-                    >>> a mismatched version in a dependency, you\s
-                    >>> must take steps to update the version of \s
-                    >>> OpenCV used in that dependency. If you do\s
-                    >>> not control that dependency and an updated\s
-                    >>> version is not available, contact the    \s
-                    >>> developers of that dependency.           \s
-                    >>>                                          \s
-                    >>> Your code will now crash.                \s
-                    >>> We hope your day gets better.            \s
-                    >>>                                          \s
-                    >>> !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\s
-                    >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\s
-                    """;
-
-            DriverStation.reportWarning(bfw, false);
-            DriverStation.reportError(bfw, false);
-            throw new UnsupportedOperationException(bfw);
-        }
         if (!Core.VERSION.equals(PhotonVersion.opencvTargetVersion)) {
             String bfw = """
 

photon-lib/src/main/java/org/photonvision/PhotonPoseEstimator.java

@@ -686,7 +686,7 @@ public class PhotonPoseEstimator {
      *
      * @param cameraResult A pipeline result from the camera.
      * @return An {@link EstimatedRobotPose} with an estimated pose, timestamp, and targets used to
-     *     create the estimate.
+     *     create the estimate, or an empty optional if there's no targets or heading data.
      */
     public Optional<EstimatedRobotPose> estimatePnpDistanceTrigSolvePose(
             PhotonPipelineResult cameraResult) {
@@ -758,7 +758,8 @@ public class PhotonPoseEstimator {
      * @param headingScaleFactor If headingFree is false, this weights the cost of changing our robot
      *     heading estimate against the tag corner reprojection error cont.
      * @return An {@link EstimatedRobotPose} with an estimated pose, timestamp, and targets used to
-     *     create the estimate.
+     *     create the estimate, or an empty optional if there's no targets or heading data, or if the
+     *     solver fails to solve the problem.
      */
     public Optional<EstimatedRobotPose> estimateConstrainedSolvepnpPose(
             PhotonPipelineResult cameraResult,
@@ -770,6 +771,18 @@ public class PhotonPoseEstimator {
         if (!shouldEstimate(cameraResult)) {
             return Optional.empty();
         }
+        // Need heading if heading fixed
+        if (!headingFree) {
+            if (headingBuffer.getSample(cameraResult.getTimestampSeconds()).isEmpty()) {
+                return Optional.empty();
+            } else {
+                // If heading fixed, force rotation component
+                seedPose =
+                        new Pose3d(
+                                seedPose.getTranslation(),
+                                new Rotation3d(headingBuffer.getSample(cameraResult.getTimestampSeconds()).get()));
+            }
+        }
         var pnpResult =
                 VisionEstimation.estimateRobotPoseConstrainedSolvepnp(
                         cameraMatrix,
@@ -799,7 +812,8 @@ public class PhotonPoseEstimator {
      *
      * @param cameraResult A pipeline result from the camera.
      * @return An {@link EstimatedRobotPose} with an estimated pose, timestamp, and targets used to
-     *     create the estimate.
+     *     create the estimate, or an empty optional if there's no targets, no multi-tag results, or
+     *     multi-tag is disabled in the web UI.
      */
     public Optional<EstimatedRobotPose> estimateCoprocMultiTagPose(
             PhotonPipelineResult cameraResult) {
@@ -829,7 +843,8 @@ public class PhotonPoseEstimator {
      * @param cameraMatrix Camera intrinsics from camera calibration data
      * @param distCoeffs Distortion coefficients from camera calibration data.
      * @return An {@link EstimatedRobotPose} with an estimated pose, timestamp, and targets used to
-     *     create the estimate.
+     *     create the estimate, or an empty optional if there's less than 2 targets visible or
+     *     SolvePnP fails.
      */
     public Optional<EstimatedRobotPose> estimateRioMultiTagPose(
             PhotonPipelineResult cameraResult, Matrix<N3, N3> cameraMatrix, Matrix<N8, N1> distCoeffs) {
@@ -861,7 +876,7 @@ public class PhotonPoseEstimator {
      *
      * @param cameraResult A pipeline result from the camera.
      * @return An {@link EstimatedRobotPose} with an estimated pose, timestamp, and targets used to
-     *     create the estimate.
+     *     create the estimate, or an empty optional if there's no targets.
      */
     public Optional<EstimatedRobotPose> estimateLowestAmbiguityPose(
             PhotonPipelineResult cameraResult) {
@@ -911,7 +926,7 @@ public class PhotonPoseEstimator {
      *
      * @param cameraResult A pipeline result from the camera.
      * @return An {@link EstimatedRobotPose} with an estimated pose, timestamp, and targets used to
-     *     create the estimate.
+     *     create the estimate, or an empty optional if there's no targets.
      */
     public Optional<EstimatedRobotPose> estimateClosestToCameraHeightPose(
             PhotonPipelineResult cameraResult) {
@@ -989,7 +1004,7 @@ public class PhotonPoseEstimator {
      * @param cameraResult A pipeline result from the camera.
      * @param referencePose reference pose to check vector magnitude difference against.
      * @return An {@link EstimatedRobotPose} with an estimated pose, timestamp, and targets used to
-     *     create the estimate.
+     *     create the estimate, or an empty optional if there's no targets.
      */
     public Optional<EstimatedRobotPose> estimateClosestToReferencePose(
             PhotonPipelineResult cameraResult, Pose3d referencePose) {
@@ -1062,7 +1077,7 @@ public class PhotonPoseEstimator {
      *
      * @param cameraResult A pipeline result from the camera.
      * @return An {@link EstimatedRobotPose} with an estimated pose, timestamp, and targets used to
-     *     create the estimate.
+     *     create the estimate, or an empty optional if there's no targets.
      */
     public Optional<EstimatedRobotPose> estimateAverageBestTargetsPose(
             PhotonPipelineResult cameraResult) {

photon-lib/src/main/java/org/photonvision/simulation/PhotonCameraSim.java

@@ -442,6 +442,10 @@ public class PhotonCameraSim implements AutoCloseable {
         Mat.zeros(videoFrameSize, CvType.CV_8UC1).assignTo(videoSimFrameRaw);
 
         for (var tgt : targets) {
+            if (detectableTgts.size() >= 50) {
+                break;
+            }
+
             // pose isn't visible, skip to next
             if (!canSeeTargetPose(cameraPose, tgt)) continue;
 

photon-lib/src/main/native/cpp/photon/PhotonCamera.cpp

@@ -29,7 +29,6 @@
 #include <string_view>
 #include <vector>
 
-#include <WPILibVersion.h>
 #include <frc/Errors.h>
 #include <frc/RobotController.h>
 #include <frc/Timer.h>
@@ -47,48 +46,6 @@ static constexpr units::second_t WARN_DEBOUNCE_SEC = 5_s;
 static constexpr units::second_t HEARTBEAT_DEBOUNCE_SEC = 500_ms;
 
 inline void verifyDependencies() {
-  if (!(std::string_view{GetWPILibVersion()} ==
-        std::string_view{photon::PhotonVersion::wpilibTargetVersion})) {
-    std::string bfw =
-        "\n\n\n\n\n"
-        ">>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\n"
-        ">>> !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n"
-        ">>>                                          \n"
-        ">>> You are running an incompatible version  \n"
-        ">>> of PhotonVision !                        \n"
-        ">>>                                          \n"
-        ">>> PhotonLib ";
-    bfw += photon::PhotonVersion::versionString;
-    bfw += " is built for WPILib ";
-    bfw += photon::PhotonVersion::wpilibTargetVersion;
-    bfw +=
-        "\n"
-        ">>> but you are using WPILib ";
-    bfw += GetWPILibVersion();
-    bfw +=
-        "\n>>>                                          \n"
-        ">>> This is neither tested nor supported.    \n"
-        ">>> You MUST update WPILib, PhotonLib, or both.\n"
-        ">>> Check `./gradlew dependencies` and ensure\n"
-        ">>> all mentions of WPILib match the version \n"
-        ">>> that PhotonLib was built for. If you find a"
-        ">>> a mismatched version in a dependency, you\n"
-        ">>> must take steps to update the version of \n"
-        ">>> WPILib used in that dependency. If you do\n"
-        ">>> not control that dependency and an updated\n"
-        ">>> version is not available, contact the    \n"
-        ">>> developers of that dependency.           \n"
-        ">>>                                          \n"
-        ">>> Your code will now crash.                \n"
-        ">>> We hope your day gets better.            \n"
-        ">>>                                          \n"
-        ">>> !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!\n"
-        ">>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\n";
-
-    FRC_ReportWarning(bfw);
-    FRC_ReportError(frc::err::Error, bfw);
-    throw new std::runtime_error(std::string{bfw});
-  }
   if (!(std::string_view{cv::getVersionString()} ==
         std::string_view{photon::PhotonVersion::opencvTargetVersion})) {
     std::string bfw =
@@ -185,8 +142,8 @@ PhotonCamera::PhotonCamera(nt::NetworkTableInstance instance,
           rootTable->GetIntegerTopic("pipelineIndexRequest").Publish()),
       pipelineIndexSub(
           rootTable->GetIntegerTopic("pipelineIndexState").Subscribe(0)),
-      ledModePub(mainTable->GetIntegerTopic("ledMode").Publish()),
-      ledModeSub(mainTable->GetIntegerTopic("ledMode").Subscribe(0)),
+      ledModePub(mainTable->GetIntegerTopic("ledModeRequest").Publish()),
+      ledModeSub(mainTable->GetIntegerTopic("ledModeState").Subscribe(0)),
       versionEntry(mainTable->GetStringTopic("version").Subscribe("")),
       cameraIntrinsicsSubscriber(
           rootTable->GetDoubleArrayTopic("cameraIntrinsics").Subscribe({})),

photon-lib/src/main/native/cpp/photon/PhotonPoseEstimator.cpp

@@ -634,13 +634,18 @@ PhotonPoseEstimator::EstimateConstrainedSolvepnpPose(
   if (!ShouldEstimate(cameraResult)) {
     return std::nullopt;
   }
+  // Need heading if heading fixed
   if (!headingFree) {
-    seedPose = frc::Pose3d{
-        seedPose.Translation(),
-        frc::Rotation3d{
-            headingBuffer.Sample(cameraResult.GetTimestamp()).value()}};
+    if (!headingBuffer.Sample(cameraResult.GetTimestamp())) {
+      return std::nullopt;
+    } else {
+      // If heading fixed, force rotation component
+      seedPose = frc::Pose3d{
+          seedPose.Translation(),
+          frc::Rotation3d{
+              headingBuffer.Sample(cameraResult.GetTimestamp()).value()}};
+    }
   }
-
   std::vector<photon::PhotonTrackedTarget> targets{
       cameraResult.GetTargets().begin(), cameraResult.GetTargets().end()};
 

photon-lib/src/main/native/cpp/photon/simulation/PhotonCameraSim.cpp

@@ -131,6 +131,10 @@ PhotonPipelineResult PhotonCameraSim::Process(
   blankFrame.assignTo(videoSimFrameRaw);
 
   for (const auto& tgt : targets) {
+    if (detectableTgts.size() >= 50) {
+      break;
+    }
+
     if (!CanSeeTargetPose(cameraPose, tgt)) {
       continue;
     }

photon-lib/src/main/native/include/photon/PhotonPoseEstimator.h

@@ -294,7 +294,7 @@ class PhotonPoseEstimator {
    *
    * @param cameraResult A pipeline result from the camera.
    * @return An EstimatedRobotPose with an estimated pose, timestamp, and
-   * targets used to create the estimate.
+   * targets used to create the estimate, or std::nullopt if there's no targets.
    */
   std::optional<EstimatedRobotPose> EstimateLowestAmbiguityPose(
       PhotonPipelineResult cameraResult);
@@ -306,7 +306,7 @@ class PhotonPoseEstimator {
    *
    * @param cameraResult A pipeline result from the camera.
    * @return An EstimatedRobotPose with an estimated pose, timestamp and
-   * targets used to create the estimate.
+   * targets used to create the estimate, or std::nullopt if there's no targets.
    */
   std::optional<EstimatedRobotPose> EstimateClosestToCameraHeightPose(
       PhotonPipelineResult cameraResult);
@@ -319,7 +319,7 @@ class PhotonPoseEstimator {
    * @param referencePose reference pose to check vector magnitude difference
    * against.
    * @return An EstimatedRobotPose with an estimated pose, timestamp, and
-   * targets used to create the estimate.
+   * targets used to create the estimate, or std::nullopt if there's no targets.
    */
   std::optional<EstimatedRobotPose> EstimateClosestToReferencePose(
       PhotonPipelineResult cameraResult, frc::Pose3d referencePose);
@@ -331,7 +331,8 @@ class PhotonPoseEstimator {
    *
    * @param cameraResult A pipeline result from the camera.
    * @return An EstimatedRobotPose with an estimated pose, timestamp, and
-   * targets used to create the estimate.
+   * targets used to create the estimate or std::nullopt if there's no targets,
+   * no multi-tag results, or multi-tag is disabled in the web UI.
    */
   std::optional<EstimatedRobotPose> EstimateCoprocMultiTagPose(
       PhotonPipelineResult cameraResult);
@@ -345,7 +346,8 @@ class PhotonPoseEstimator {
    * @param cameraMatrix Camera intrinsics from camera calibration data.
    * @param distCoeffs Distortion coefficients from camera calibration data.
    * @return An EstimatedRobotPose with an estimated pose, timestamp, and
-   * targets used to create the estimate.
+   * targets used to create the estimate, or std::nullopt if there's less than 2
+   * targets visible or SolvePnP fails.
    */
   std::optional<EstimatedRobotPose> EstimateRioMultiTagPose(
       PhotonPipelineResult cameraResult, PhotonCamera::CameraMatrix camMat,
@@ -363,7 +365,8 @@ class PhotonPoseEstimator {
    *
    * @param cameraResult A pipeline result from the camera.
    * @return An EstimatedRobotPose with an estimated pose, timestamp, and
-   * targets used to create the estimate.
+   * targets used to create the estimate, or std::nullopt if there's no targets
+   * or heading data.
    */
   std::optional<EstimatedRobotPose> EstimatePnpDistanceTrigSolvePose(
       PhotonPipelineResult cameraResult);
@@ -372,7 +375,7 @@ class PhotonPoseEstimator {
    * Return the average of the best target poses using ambiguity as weight.
    * @param cameraResult A pipeline result from the camera.
    * @return An EstimatedRobotPose with an estimated pose, timestamp, and
-   * targets used to create the estimate.
+   * targets used to create the estimate, or std::nullopt if there's no targets.
    */
   std::optional<EstimatedRobotPose> EstimateAverageBestTargetsPose(
       PhotonPipelineResult cameraResult);
@@ -401,7 +404,8 @@ class PhotonPoseEstimator {
    * changing our robot heading estimate against the tag corner reprojection
    * error cost.
    * @return An EstimatedRobotPose with an estimated pose, timestamp, and
-   * targets used to create the estimate.
+   * targets used to create the estimate, or std::nullopt if there's no targets
+   * or heading data, or if the solver fails to solve the problem.
    */
   std::optional<EstimatedRobotPose> EstimateConstrainedSolvepnpPose(
       photon::PhotonPipelineResult cameraResult,

photon-lib/src/test/java/org/photonvision/VisionSystemSimTest.java

@@ -179,6 +179,28 @@ class VisionSystemSimTest {
         assertTrue(result.hasTargets());
     }
 
+    @Test
+    public void testBunchaTargets() {
+        var visionSysSim = new VisionSystemSim("Test");
+        var camera = new PhotonCamera(inst, "camera");
+        var cameraSim = new PhotonCameraSim(camera);
+        visionSysSim.addCamera(cameraSim, new Transform3d());
+        cameraSim.prop.setCalibration(640, 480, Rotation2d.fromDegrees(80));
+
+        for (int i = 0; i < 100; i++) {
+            final var targetPose =
+                    new Pose3d(new Translation3d(15.98 + i * 0.1, 0, 1), new Rotation3d(0, 0, Math.PI));
+            visionSysSim.addVisionTargets(new VisionTargetSim(targetPose, new TargetModel(0.5, 0.5), i));
+        }
+
+        var robotPose = new Pose2d(new Translation2d(5, 0), Rotation2d.fromDegrees(5));
+        visionSysSim.update(robotPose);
+
+        var res = waitForSequenceNumber(camera, 1);
+
+        assertEquals(50, res.getTargets().size());
+    }
+
     @Test
     public void testNotVisibleVert1() {
         final var targetPose =

photon-lib/src/test/native/cpp/VisionSystemSimTest.cpp

@@ -117,6 +117,29 @@ TEST_F(VisionSystemSimTest, TestVisibilityCupidShuffle) {
   ASSERT_TRUE(camera.GetLatestResult().HasTargets());
 }
 
+TEST_F(VisionSystemSimTest, TestBunchaTargets) {
+  photon::VisionSystemSim visionSysSim{"Test"};
+  photon::PhotonCamera camera{"camera"};
+  photon::PhotonCameraSim cameraSim{&camera};
+  visionSysSim.AddCamera(&cameraSim, frc::Transform3d{});
+  cameraSim.prop.SetCalibration(640, 480, frc::Rotation2d{80_deg});
+
+  std::vector<photon::VisionTargetSim> targets;
+  for (int i = 0; i < 100; i++) {
+    targets.emplace_back(
+        frc::Pose3d{
+            frc::Translation3d{15.98_m + i * 0.1_m, 0_m, 1_m},
+            frc::Rotation3d{0_rad, 0_rad, units::radian_t{std::numbers::pi}}},
+        photon::TargetModel{0.5_m, 0.5_m}, i);
+  }
+  visionSysSim.AddVisionTargets(targets);
+
+  frc::Pose2d robotPose{frc::Translation2d{5_m, 0_m}, frc::Rotation2d{5_deg}};
+  visionSysSim.Update(robotPose);
+
+  ASSERT_EQ(camera.GetLatestResult().targets.size(), 50u);
+}
+
 TEST_F(VisionSystemSimTest, TestNotVisibleVert1) {
   frc::Pose3d targetPose{
       frc::Translation3d{15.98_m, 0_m, 1_m},

photon-server/src/main/java/org/photonvision/server/RequestHandler.java

@@ -39,7 +39,7 @@ import org.opencv.imgcodecs.Imgcodecs;
 import org.photonvision.common.configuration.ConfigManager;
 import org.photonvision.common.configuration.NetworkConfig;
 import org.photonvision.common.configuration.NeuralNetworkModelManager;
-import org.photonvision.common.configuration.NeuralNetworkPropertyManager.ModelProperties;
+import org.photonvision.common.configuration.NeuralNetworkModelsSettings.ModelProperties;
 import org.photonvision.common.dataflow.DataChangeDestination;
 import org.photonvision.common.dataflow.DataChangeService;
 import org.photonvision.common.dataflow.events.IncomingWebSocketEvent;
@@ -852,6 +852,12 @@ public class RequestHandler {
             ctx.result("There was an error while saving the uploaded object detection models");
             logger.error("There was an error while saving the uploaded object detection models");
         }
+
+        DataChangeService.getInstance()
+                .publishEvent(
+                        new OutgoingUIEvent<>(
+                                "fullsettings",
+                                UIPhotonConfiguration.programStateToUi(ConfigManager.getInstance().getConfig())));
     }
 
     private record DeleteObjectDetectionModelRequest(Path modelPath) {}
@@ -898,17 +904,17 @@ public class RequestHandler {
 
             ctx.status(200).result("Successfully deleted object detection model");
 
+            DataChangeService.getInstance()
+                    .publishEvent(
+                            new OutgoingUIEvent<>(
+                                    "fullsettings",
+                                    UIPhotonConfiguration.programStateToUi(ConfigManager.getInstance().getConfig())));
+
         } catch (Exception e) {
             ctx.status(500);
             ctx.result("Error deleting object detection model: " + e.getMessage());
             logger.error("Error deleting object detection model", e);
         }
-
-        DataChangeService.getInstance()
-                .publishEvent(
-                        new OutgoingUIEvent<>(
-                                "fullsettings",
-                                UIPhotonConfiguration.programStateToUi(ConfigManager.getInstance().getConfig())));
     }
 
     private record RenameObjectDetectionModelRequest(Path modelPath, String newName) {}
@@ -951,6 +957,12 @@ public class RequestHandler {
 
             NeuralNetworkModelManager.getInstance().discoverModels();
             ctx.status(200).result("Successfully renamed object detection model");
+
+            DataChangeService.getInstance()
+                    .publishEvent(
+                            new OutgoingUIEvent<>(
+                                    "fullsettings",
+                                    UIPhotonConfiguration.programStateToUi(ConfigManager.getInstance().getConfig())));
         } catch (Exception e) {
             ctx.status(500);
             ctx.result("Error renaming object detection model: " + e.getMessage());
@@ -970,6 +982,12 @@ public class RequestHandler {
             ctx.result("Error clearing object detection models: " + e.getMessage());
             logger.error("Error clearing object detection models", e);
         }
+
+        DataChangeService.getInstance()
+                .publishEvent(
+                        new OutgoingUIEvent<>(
+                                "fullsettings",
+                                UIPhotonConfiguration.programStateToUi(ConfigManager.getInstance().getConfig())));
     }
 
     public static void onDeviceRestartRequest(Context ctx) {

photon-targeting/src/main/java/org/photonvision/common/dataflow/structures/Packet.java

@@ -31,6 +31,8 @@ public class Packet {
     // Read and write positions.
     int readPos, writePos;
 
+    public static final int MAX_ARRAY_LEN = Byte.MAX_VALUE;
+
     /**
      * Constructs an empty packet. This buffer will dynamically expand if we need more data space.
      *

photonlib-cpp-examples/aimandrange/build.gradle

@@ -1,7 +1,7 @@
 plugins {
     id "cpp"
     id "google-test-test-suite"
-    id "edu.wpi.first.GradleRIO" version "2026.1.1"
+    id "edu.wpi.first.GradleRIO" version "2026.2.1"
 }
 
 repositories {
@@ -11,8 +11,8 @@ repositories {
 
 wpi.maven.useLocal = false
 wpi.maven.useDevelopment = false
-wpi.versions.wpilibVersion = "2026.1.1"
-wpi.versions.wpimathVersion = "2026.1.1"
+wpi.versions.wpilibVersion = "2026.2.1"
+wpi.versions.wpimathVersion = "2026.2.1"
 
 // Define my targets (RoboRIO) and artifacts (deployable files)
 // This is added by GradleRIO's backing project DeployUtils.

photonlib-cpp-examples/aimattarget/build.gradle

@@ -1,7 +1,7 @@
 plugins {
     id "cpp"
     id "google-test-test-suite"
-    id "edu.wpi.first.GradleRIO" version "2026.1.1"
+    id "edu.wpi.first.GradleRIO" version "2026.2.1"
 }
 
 repositories {
@@ -11,8 +11,8 @@ repositories {
 
 wpi.maven.useLocal = false
 wpi.maven.useDevelopment = false
-wpi.versions.wpilibVersion = "2026.1.1"
-wpi.versions.wpimathVersion = "2026.1.1"
+wpi.versions.wpilibVersion = "2026.2.1"
+wpi.versions.wpimathVersion = "2026.2.1"
 
 // Define my targets (RoboRIO) and artifacts (deployable files)
 // This is added by GradleRIO's backing project DeployUtils.

photonlib-cpp-examples/poseest/build.gradle

@@ -1,7 +1,7 @@
 plugins {
     id "cpp"
     id "google-test-test-suite"
-    id "edu.wpi.first.GradleRIO" version "2026.1.1"
+    id "edu.wpi.first.GradleRIO" version "2026.2.1"
 }
 
 repositories {
@@ -11,8 +11,8 @@ repositories {
 
 wpi.maven.useLocal = false
 wpi.maven.useDevelopment = false
-wpi.versions.wpilibVersion = "2026.1.1"
-wpi.versions.wpimathVersion = "2026.1.1"
+wpi.versions.wpilibVersion = "2026.2.1"
+wpi.versions.wpimathVersion = "2026.2.1"
 
 // Define my targets (RoboRIO) and artifacts (deployable files)
 // This is added by GradleRIO's backing project DeployUtils.

photonlib-java-examples/aimandrange/build.gradle

@@ -1,6 +1,6 @@
 plugins {
     id "java"
-    id "edu.wpi.first.GradleRIO" version "2026.1.1"
+    id "edu.wpi.first.GradleRIO" version "2026.2.1"
 }
 
 sourceCompatibility = JavaVersion.VERSION_17
@@ -13,8 +13,8 @@ repositories {
 }
 
 wpi.maven.useDevelopment = true
-wpi.versions.wpilibVersion = "2026.1.1"
-wpi.versions.wpimathVersion = "2026.1.1"
+wpi.versions.wpilibVersion = "2026.2.1"
+wpi.versions.wpimathVersion = "2026.2.1"
 
 
 // Define my targets (RoboRIO) and artifacts (deployable files)

photonlib-java-examples/aimattarget/build.gradle

@@ -1,6 +1,6 @@
 plugins {
     id "java"
-    id "edu.wpi.first.GradleRIO" version "2026.1.1"
+    id "edu.wpi.first.GradleRIO" version "2026.2.1"
 }
 
 sourceCompatibility = JavaVersion.VERSION_17
@@ -9,8 +9,8 @@ targetCompatibility = JavaVersion.VERSION_17
 def ROBOT_MAIN_CLASS = "frc.robot.Main"
 
 wpi.maven.useDevelopment = true
-wpi.versions.wpilibVersion = "2026.1.1"
-wpi.versions.wpimathVersion = "2026.1.1"
+wpi.versions.wpilibVersion = "2026.2.1"
+wpi.versions.wpimathVersion = "2026.2.1"
 
 
 // Define my targets (RoboRIO) and artifacts (deployable files)

photonlib-java-examples/poseest/build.gradle

@@ -1,6 +1,6 @@
 plugins {
     id "java"
-    id "edu.wpi.first.GradleRIO" version "2026.1.1"
+    id "edu.wpi.first.GradleRIO" version "2026.2.1"
 }
 
 sourceCompatibility = JavaVersion.VERSION_17
@@ -9,8 +9,8 @@ targetCompatibility = JavaVersion.VERSION_17
 def ROBOT_MAIN_CLASS = "frc.robot.Main"
 
 wpi.maven.useDevelopment = true
-wpi.versions.wpilibVersion = "2026.1.1"
-wpi.versions.wpimathVersion = "2026.1.1"
+wpi.versions.wpilibVersion = "2026.2.1"
+wpi.versions.wpimathVersion = "2026.2.1"
 
 
 // Define my targets (RoboRIO) and artifacts (deployable files)

photonlib-python-examples/aimandrange/pyproject.toml

@@ -6,7 +6,7 @@
 [tool.robotpy]
 
 # Version of robotpy this project depends on
-robotpy_version = "2026.1.1"
+robotpy_version = "2026.2.1"
 
 # Which extra RobotPy components should be installed
 # -> equivalent to `pip install robotpy[extra1, ...]

photonlib-python-examples/aimattarget/pyproject.toml

@@ -6,7 +6,7 @@
 [tool.robotpy]
 
 # Version of robotpy this project depends on
-robotpy_version = "2026.1.1"
+robotpy_version = "2026.2.1"
 
 # Which extra RobotPy components should be installed
 # -> equivalent to `pip install robotpy[extra1, ...]

photonlib-python-examples/poseest/pyproject.toml

@@ -6,7 +6,7 @@
 [tool.robotpy]
 
 # Version of robotpy this project depends on
-robotpy_version = "2026.1.1"
+robotpy_version = "2026.2.1"
 
 # Which extra RobotPy components should be installed
 # -> equivalent to `pip install robotpy[extra1, ...]

photonlib-python-examples/run.sh

@@ -5,19 +5,6 @@ if [ $# -eq 0 ]
     exit 1
 fi
 
-# To run any example, we want to use photonlib out of this repo
-# Build the wheel first
-pushd ../photon-lib/py
-if [ -d build ]
-  then rm -rdf build
-fi
-python3 setup.py bdist_wheel
-popd
-
-# Add the output directory to PYTHONPATH to make sure it gets picked up
-export PHOTONLIBPY_ROOT=../photon-lib/py
-export PYTHONPATH=$PHOTONLIBPY_ROOT
-
 # Move to the right example folder
 cd $1
 

scripts/rknn_conversion.ipynb

@@ -33,6 +33,27 @@
    ],
    "id": "500d656b7cc0ebd7"
   },
+  {
+   "metadata": {},
+   "cell_type": "markdown",
+   "source": [
+    "#### *Numpy Fix* - Important for Google Colab Users\n",
+    "\n",
+    "Google Colab comes with an incompatible version of Numpy installed. To fix this, please run the following cells below and **restart your session** when prompted."
+   ],
+   "id": "b3a9e1a334bce144"
+  },
+  {
+   "metadata": {},
+   "cell_type": "code",
+   "outputs": [],
+   "execution_count": null,
+   "source": [
+    "%pip uninstall numpy -y\n",
+    "%pip install \"numpy==1.26.4\""
+   ],
+   "id": "7156e69495f48f49"
+  },
   {
    "metadata": {},
    "cell_type": "markdown",
@@ -41,7 +62,7 @@
     "\n",
     "Please run the cell below to be able to use the `create_onnx` and `create_rknn` functions."
    ],
-   "id": "798298b1dbe33d2d"
+   "id": "51566ff74470e57"
   },
   {
    "metadata": {},
@@ -132,6 +153,7 @@
     "            os.path.join(ultralytics_folder_name_yolov5, \"requirements.txt\"),\n",
     "            \"torch<2.6.0\",\n",
     "            \"onnx==1.18.0\",\n",
+    "            \"numpy==1.26.4\",\n",
     "            \"onnxscript\",\n",
     "        ]\n",
     "    )\n",
@@ -172,7 +194,7 @@
     "def run_onnx_conversion_no_anchor(model_path):\n",
     "    check_or_clone_rockchip_repo(yolo_non_anchor_repo)\n",
     "    run_pip_install_or_else_exit(\n",
-    "        [\"-e\", ultralytics_default_folder_name, \"onnx==1.18.0\", \"onnxscript\"]\n",
+    "        [\"-e\", ultralytics_default_folder_name, \"onnx==1.18.0\", \"numpy==1.26.4\", \"onnxscript\"]\n",
     "    )\n",
     "\n",
     "    sys.path.insert(0, os.path.abspath(ultralytics_default_folder_name))\n",
@@ -384,28 +406,7 @@
     "    except SystemExit:\n",
     "        print(\"RKNN Conversion failed, see output above\")\n"
    ],
-   "id": "ea6869140a61126d"
-  },
-  {
-   "metadata": {},
-   "cell_type": "markdown",
-   "source": [
-    "#### *Numpy Fix* - Important for Google Colab Users\n",
-    "\n",
-    "Google Colab comes with an incompatible version of Numpy installed. To fix this, please run the following cells below and **restart your session** when prompted."
-   ],
-   "id": "b3a9e1a334bce144"
-  },
-  {
-   "metadata": {},
-   "cell_type": "code",
-   "outputs": [],
-   "execution_count": null,
-   "source": [
-    "%pip uninstall numpy -y\n",
-    "%pip install \"numpy>=1.23.0,<2.0.0\""
-   ],
-   "id": "7156e69495f48f49"
+   "id": "4d7a8adee7a03377"
   },
   {
    "metadata": {},