All Classes and Interfaces
Class
Description
Detector parameters.
Represents a pipeline that blurs the image.
What kind of camera lens model our intrinsics are modeling.
Holds various helper geometries describing the relation between camera and target.
Represents a pipe that collects available 2d targets.
Determines the target corners of the
TrackedTarget.Defines a pipe.
Add migrations by adding the SQL commands for each migration sequentially to this array.
An estimated pose based on pipeline result
A
FrameProvider that will read and provide an image from a path.Represents the properties of a frame.
An image rotation about the camera's +Z axis, which points out of the camera towards the world.
JSON-serializable image.
JSON-serializable image.
Listens for and reproduces Linux kernel logs, from /var/log/kern.log, into the Photon logger
ecosystem
On the OV5649 the actual FPS we want to request from the GPU can be higher than the FPS that we
can do after processing.
Helper for extracting photon-libcamera-gl-driver shared library files.
TODO: get rid of static {} blocks and refactor to singleton pattern
Estimate the camera pose given multiple Apriltag observations
Auto-generated serialization/deserialization helper for MultiTargetPNPResult
Manages the loading of neural network models.
This class is a wrapper around all per-pipeline NT topics that PhotonVision should be publishing
It's split here so the sim and real-camera implementations can share a common implementation of
the naming and registration of the NT content.
A 'null' implementation of the
Model and ObjectDetector interfaces.ObjectDetector lifecycle:
Models are discovered by NeuralNetworkModelManager
Model is selected as a parameter of ObjectDetectionPipe
Model.load() is called to create a ObjectDetector instance
ObjectDetector.detect(Mat, double, double) is called to perform object detection
Releasable.release() is called to release resources
Our blessed images inject the current version via this build workflow:
https://github.com/PhotonVision/photon-image-modifier/blob/2e5ddb6b599df0be921c12c8dbe7b939ecd7f615/.github/workflows/main.yml#L67
This is a "fake" pipeline that is just used to move identical pipe sets out of real pipelines.
A packet that holds byte-packed data to be sent over NetworkTables.
Protobuf type
ProtobufMultiTargetPNPResultProtobuf type
ProtobufPhotonPipelineResultProtobuf type
ProtobufPhotonTrackedTargetProtobuf type
ProtobufPNPResultProtobuf type
ProtobufTargetCornerThis class wraps an
ArucoDetector for convenience.Represents a camera that is connected to PhotonVision.
A handle for simulating
PhotonCamera values.Auto-generated serialization/deserialization helper for PhotonPipelineMetadata
Represents a pipeline result from a PhotonCamera.
Auto-generated serialization/deserialization helper for PhotonPipelineResult
The PhotonPoseEstimator class filters or combines readings from all the AprilTags visible at a
given timestamp on the field to produce a single robot in field pose, using the strategy set
below.
Position estimation strategies that can be used by the
PhotonPoseEstimator class.Auto-generated serialization/deserialization helper for PhotonTrackedTarget
A class that defines the exceptions that can be thrown by Pigpio.
The best estimated transformation from solvePnP, and possibly an alternate transformation
depending on the solvePNP method.
Auto-generated serialization/deserialization helper for PnpResult
Redirect cscore logs to our logger
Represents a pipeline for tracking retro-reflective targets.
Pipe that resizes an image to a given resolution
Manages an object detector using the rknn backend.
Pipe that rotates an image to a given orientation
Represents a transformation that first rotates a pose around the origin, and then translates it.
Execute external process and optionally read output buffer.
Calibration and performance values for this camera.
Saves settings in a SQLite database file (called photon.sqlite).
Represents a point in an image at the corner of the minimum-area bounding rectangle, in pixels.
Auto-generated serialization/deserialization helper for TargetCorner
Describes the 3d model of a target.
A model representing the vertices of targets with known shapes.
Dummy class for unit testing the vision source manager
Send ping-pongs to estimate server time, relative to nt::Now.
Helper to hold a single TimeSyncServer instance with some default config
This is the God Class
VisionModuleManager has many VisionModules, and provides camera configuration data to them.
VisionRunner has a frame supplier, a pipeline supplier, and a result consumer
A simulated vision system involving a camera(s) and coprocessor(s) mounted on a mobile robot
running PhotonVision, detecting targets placed on the field.
Describes a vision target located somewhere on the field that your vision system can detect.