robokudo.annotators.byte_track_annotator

Classes

ByteTrackAnnotator

A ByteTrack annotator that tracks detected objects across pipeline ticks.

Module Contents

class robokudo.annotators.byte_track_annotator.ByteTrackAnnotator(name: str = 'ByteTrackAnnotator')

Bases: robokudo.annotators.core.BaseAnnotator

A ByteTrack annotator that tracks detected objects across pipeline ticks.

The annotator is capable of utilizing:

  • ObjectHypothesis ROI

  • ObjectHypothesis ROI + Mask

  • ObjectHypothesis ROI + Mask + Classification annotation

tracker: supervision.ByteTrack

Main ByteTrack instance for object instance tracking.

last_ts: float = 0.0

Timestamp of the last data received, used for tracker reset detection.

label_annotator

A special label annotator, that cannot simply be used like the annotators in self.vis_annotators.

vis_annotators

A list of simple supervision annotators to use for visualization. They will be applied in order.

update() py_trees.common.Status
visualize(tracked_detections: supervision.Detections) None

Visualize the tracked detections and set the resulting image in the annotator output struct.

Parameters:

tracked_detections – The tracked detections to visualize.