Point Cloud Bounding Box to 2D Bounding Box
Function: Convert the point cloud bounding box to a 2d pixel bounding box by internal parameters.
Input Parameters:
| Name | Type | Valid Range | Default Value | Meaning |
|---|---|---|---|---|
Point Cloud Bounding Box |
Box3D |
None |
None |
List of camera coordinate system point cloud point cloud bounding boxes in the format [{'center': [center_x, center_y, center_z], 'size': [size_x, size_y, size_z], 'transform': [3 * 3]}]. |
Camera internal reference |
Matrix |
None |
None |
In-camera reference matrix. |
Image |
ColorImage |
None |
None |
Input image, need RGB channel color image, for drawing, can input the original image, it is recommended to input the display image output by the deep learning node, in order to better view the effect after drawing, is optional input. |
Output Parameters:
| Name | Type | Valid Range | Default Value | Meaning |
|---|---|---|---|---|
Pixel bounding box |
Box2D |
None |
None |
Pixel bounding box list, outputs the 2d pixel bounding box corresponding to the point cloud bounding box, in the format Nx4x2, where N represents N pixel bounding boxes, and 4x2 represents the four corners of each bounding box order. |
2D ROI |
ROI2D |
None |
None |
2D ROI in the format [{'min': [min_x, min_y], 'max': [max_x, max_y]}]. |