Using WPILib Pose Estimation, Simulation, and PhotonVision Together

The following example comes from the PhotonLib example repository (Java). Full code is available at that links.

Knowledge and Equipment Needed

Background

This example builds upon WPILib’s Differential Drive Pose Estimator. It adds a PhotonCamera to gather estimates of the robot’s position on the field. This in turn can be used for aligning with vision targets, and increasing accuracy of autonomous routines.

To support simulation, a SimVisionSystem is used to drive data into the PhotonCamera. The far high goal target from 2020 is modeled.

Walkthrough

WPILib’s Pose2d class is used to represent robot positions on the field.

Three different Pose2d positions are relevant for this example:

  1. Desired Pose: The location the some autonomous routine wants the robot to be in.

  2. Estimated Pose: The location the software believes the robot to be in, based on physics models and sensor feedback.

  3. Actual Pose: The locations he robot is actually at. The physics simulation generates this in simulation, but it cannot be directly measured on the real robot.

Estimating Pose

The DrivetrainPoseEstimator class is responsible for generating an estimated robot pose using sensor readings (including PhotonVision).

Please reference the WPILib documentation on using the DifferentialDrivePoseEstimator class.

For both simulation and on-robot code, we create objects to represent the physical location and size of the vision targets we are calibrated to detect. This example models the down-field high goal vision target from the 2020 and 2021 games.

 79    // See
 80    // https://firstfrc.blob.core.windows.net/frc2020/PlayingField/2020FieldDrawing-SeasonSpecific.pdf
 81    // page 208
 82    public static final double targetWidth =
 83            Units.inchesToMeters(41.30) - Units.inchesToMeters(6.70); // meters
 84
 85    // See
 86    // https://firstfrc.blob.core.windows.net/frc2020/PlayingField/2020FieldDrawing-SeasonSpecific.pdf
 87    // page 197
 88    public static final double targetHeight =
 89            Units.inchesToMeters(98.19) - Units.inchesToMeters(81.19); // meters
 90    public static final double targetHeightAboveGround = Units.inchesToMeters(81.19); // meters
 91
 92    // See https://firstfrc.blob.core.windows.net/frc2020/PlayingField/LayoutandMarkingDiagram.pdf
 93    // pages 4 and 5
 94    public static final double kFarTgtXPos = Units.feetToMeters(54);
 95    public static final double kFarTgtYPos =
 96            Units.feetToMeters(27 / 2) - Units.inchesToMeters(43.75) - Units.inchesToMeters(48.0 / 2.0);
 97    public static final Pose2d kFarTargetPose =
 98            new Pose2d(new Translation2d(kFarTgtXPos, kFarTgtYPos), new Rotation2d(0.0));
 99
100    public static final SimVisionTarget kFarTarget =
101            new SimVisionTarget(kFarTargetPose, targetHeightAboveGround, targetWidth, targetHeight);

//Coming soon!

To incorporate Photon Vision, we need to create a PhotonCamera:

49    private PhotonCamera cam = new PhotonCamera(Constants.kCamName);

//Coming soon!

During periodic execution, we read back camera results. If we see a target in the image, we pass the camera-measured pose of the robot to the DifferentialDrivePoseEstimator.

82    public void update(
83            DifferentialDriveWheelSpeeds actWheelSpeeds, double leftDist, double rightDist) {
84        m_poseEstimator.update(gyro.getRotation2d(), actWheelSpeeds, leftDist, rightDist);
85
86        var res = cam.getLatestResult();
87        if (res.hasTargets()) {
88            double imageCaptureTime = Timer.getFPGATimestamp() - res.getLatencyMillis();
89            Transform2d camToTargetTrans = res.getBestTarget().getCameraToTarget();
90            Pose2d camPose = Constants.kFarTargetPose.transformBy(camToTargetTrans.inverse());
91            m_poseEstimator.addVisionMeasurement(
92                    camPose.transformBy(Constants.kCameraToRobot), imageCaptureTime);
93        }
94    }

//Coming soon!

That’s it!

Simulating the Camera

First, we create a new SimVisionSystem to represent our camera and coprocessor running PhotonVision.

77    // Simulated Vision System.
78    // Configure these to match your PhotonVision Camera,
79    // pipeline, and LED setup.
80    double camDiagFOV = 75.0; // degrees
81    double camPitch = 15.0; // degrees
82    double camHeightOffGround = 0.85; // meters
83    double maxLEDRange = 20; // meters
84    int camResolutionWidth = 640; // pixels
85    int camResolutionHeight = 480; // pixels
86    double minTargetArea = 10; // square pixels
87
88    SimVisionSystem simVision =
89            new SimVisionSystem(
90                    Constants.kCamName,
91                    camDiagFOV,
92                    camPitch,
93                    Constants.kCameraToRobot,
94                    camHeightOffGround,
95                    maxLEDRange,
96                    camResolutionWidth,
97                    camResolutionHeight,
98                    minTargetArea);

//Coming soon!

Then, we add our target to the simulated vision system.

100    public DrivetrainSim() {
101        simVision.addSimVisionTarget(Constants.kFarTarget);
102    }

//Coming soon!

If you have additional targets you want to detect, you can add them in the same way as the first one.

Updating the Simulated Vision System

Once we have all the properties of our simulated vision system defined, the remaining work is minimal. Periodically, pass in the robot’s pose to the simulated vision system.

141        // Update PhotonVision based on our new robot position.
142        simVision.processFrame(drivetrainSimulator.getPose());

//Coming soon!

The rest is done behind the scenes.