Skip to main content

Camera to robot pose

CameraToRobotPose converts a field-relative camera pose into a field-relative robot pose using the selected camera's extrinsics.

Inputs​

camera_pose is a finite 4 by 4 transformation matrix, or None.

Outputs​

robot_pose is a 4 by 4 transformation matrix. Invalid matrices and None produce None.

When to use​

Use this after camera localization when the pipeline needs the robot origin rather than the camera origin.

Configuration​

  • camera_bus_id: Required bus ID used to read camera pitch, yaw, roll, and XYZ offsets.

Limitations​

The operation assumes project coordinate conventions. If no camera configuration registry is available, it uses identity extrinsics.