Open CBY-9527 opened 1 year ago
The depth map is only generated from the center patch (352x1216) of an RGB image due to the different sizes of RGB images in the KITTI dataset. When we gather RGB information with a depth map, we will pad the depth map to the size of the corresponding RGB image. Please refer to this line.
Excuse me, and I have a question. When I used depth_to_lidar.py to convert the depth map to a pseudo point cloud, I found that the size of the depth map (352x1216) is different from the size of the RGB image (375x1242). The features of the pseudo point cloud generated by the code are composed of x, y, z, r, g, b, seg, u, and v. Will the inconsistent depth and RGB image size cause the obtained R, G, B values of the U, V position do not correspond to the real R, G, B values?