vision_detect_plugvision_detect_plug action is used to detect the pose of a plug in a gripper.
Action Goalvision_detect_plug/goal (pr2_plugs_msgs/VisionPlugDetectionGoal)
- The goal gives the camera to use, a prior pose of an outlet and a wall norm.
Action Resultvision_detect_plug/result (pr2_plugs_msgs/VisionPlugDetectionResult)
- This returns the pose of the outlet.
Subscribed Topics<camera_name>/image_rect (sensor_msgs/Image)
- This subscribes to the image from the camera.
Published Topics~display_image (sensor_msgs/Image)
- Camera image marked with detected checkerboard corners.
Checkerboard dimensionsThe checkerboard detector looks for the interior corners; for example, a standard checkerboard is considered to have dimensions 7x7.
- The number of interior corners in the horizontal direction.
- The number of interior corners in the vertical direction.
- The length of a checkerboard square edge (meters).
- Whether to perform subpixel refinement on detected corners. This usually yields a more accurate pose.
Checkerboard to plug transformplug_position_* is the position of the plug frame origin in the checkerboard frame. plug_orientation_* is a quaternion rotating the plug frame to the checkerboard frame. With the plug positioned for entering the outlet, the checkerboard frame is:
- Origin: The lower-right interior checkerboard corner.
- X: Down
- Y: Forward (towards the outlet)
- Z: Into the plug
- X coordinate.
- Y coordinate.
- Z coordinate.
- Quaternion X.
- Quaternion Y.
- Quaternion Z.
- Quaternion W.