How the point cloud semantically segmented file is encoded？
magical1998 opened this issue · 1 comments
magical1998 commented
What is the specific encoding method of the point cloud segmentation file, and is it suitable for the first 16 bits of SenmanticKITTI for semantic segmentation, and the last 16 bits for panoramic segmentation?
whyekit-motional commented
The labels for each point in a given lidar point cloud is simply a numpy array which is saved into a .bin file via numpy's tofile(...): https://numpy.org/doc/stable/reference/generated/numpy.ndarray.tofile.html