rect_array_in_panorama_to_bounding_box_array.py¶

jsk_perception/nodes/images/rect_array_in_panorama_to_bounding_box_array.gif

This node calculates distances to RectArray in a panorama image with FOV of the image, heights of rects and heights of each labels with the equation (object height [m]) = (distance [m]) * (object height in panorama image [rad]), and publish BoundingBoxArray.

Sample is sample_rect_array_in_panorama_to_bounding_box_array.launch

Subscribing Topics¶

~panorama_info (jsk_recognition_msgs/PanoramaInfo)

Panorama info topic.

~input_class (jsk_recognition_msgs/ClassificationResult)

ClassificationResult from object detection with the panorama image topic

~input_rects (jsk_recognition_msgs/RectArray)

RectArray from object detection with the panorama image topic

Publishing Topics¶

~bbox_array (jsk_recognition_msgs/BoundingBoxArray)

BoundingBoxArray of each object of output of object detection.

Parameters¶

~frame_fixed (string, default: fixed_frame)

The frame_id of BoundingBoxArray. Assumed to be a fixed frame.

~dimensions_labels (dict of string to lists of float values, default: {})

Dimensions for each labels. These are used for BoundingBoxArray and also calculation of distance.

e.g. {'person':[0.5,0.5,1.5]}

~duration_timeout (float, default: `0.05)

Duration of timeout for lookup transform

Sample¶

roslaunch jsk_perception sample_rect_array_in_panorama_to_bounding_box_array.launch