robokudo.annotators.size_bb

Classes

SizeBBAnnotator

A class to analyze 3D point clouds and compute oriented bounding box sizes.

Module Contents

class robokudo.annotators.size_bb.SizeBBAnnotator(name='SizeBBAnnotator')

Bases: robokudo.annotators.core.BaseAnnotator

A class to analyze 3D point clouds and compute oriented bounding box sizes.

This annotator processes object hypotheses by computing their oriented 3D bounding boxes and storing size information as annotations. It performs coordinate transformations and uses minimum area rectangles to determine object dimensions.

Parameters:

name (str) – Name of the annotator instance

update() → py_trees.common.Status

Process object hypotheses to compute and annotate their 3D bounding box sizes.

For each object hypothesis with sufficient points:

  • Transforms point cloud to world coordinates

  • Projects points to 2D and finds minimum area rectangle

  • Computes 3D oriented bounding box

  • Annotates size information

  • Updates visualization

Returns:

SUCCESS if processing completed, FAILURE if camera viewpoint not found

Return type:

py_trees.Status

Raises:

Exception – If camera viewpoint transform cannot be found in CAS