robokudo.annotators.pointcloud_crop_map¶

Point cloud cropping in world coordinates.

This module provides an annotator for:

Cropping point clouds in world coordinate frame
Handling camera-to-world transformations
Supporting consistent spatial segmentation
Transforming results back to camera frame

Warning

This annotator is deprecated. Use pointcloud_crop with relative_to_world=True instead.

Classes¶

PointcloudCropMapAnnotator

Point cloud cropping in world coordinates.

Module Contents¶

class robokudo.annotators.pointcloud_crop_map.PointcloudCropMapAnnotator(name='PointcloudCropMapAnnotator', descriptor=Descriptor())¶

Bases: robokudo.annotators.core.BaseAnnotator

Point cloud cropping in world coordinates.

This annotator:

Transforms point cloud to world frame
Crops using axis-aligned bounding box
Transforms cropped cloud back to camera frame
Segments consistent world regions

Warning

Deprecated: Use pointcloud_crop with relative_to_world=True instead.

class Descriptor¶

Bases: robokudo.annotators.core.BaseAnnotator.Descriptor

Configuration descriptor for world-frame point cloud cropping.

class Parameters¶

Parameters for configuring world-frame cropping.

Bounding box parameters in world coordinates:

Variables:

min_x – Minimum X coordinate, defaults to -1.5
min_y – Minimum Y coordinate, defaults to 0.0
min_z – Minimum Z coordinate, defaults to -9.0
max_x – Maximum X coordinate, defaults to 1.0
max_y – Maximum Y coordinate, defaults to 4.0
max_z – Maximum Z coordinate, defaults to 3.0

min_x = -1.5¶

min_y = -0.0¶

min_z = -9.0¶

max_x = 1.0¶

max_y = 4.0¶

max_z = 3.0¶

parameters¶

color = None¶

update()¶

Process and crop point cloud in world coordinates.

The method:

Gets camera-to-world transform from CAS
Transforms point cloud to world frame
Crops using axis-aligned bounding box
Transforms cropped cloud back to camera frame
Updates CAS with cropped cloud

Returns:: SUCCESS after processing, FAILURE if transform not found
Return type:: py_trees.Status