depth_to_pointcloud_python

I couldn't find a Python implementation of ROS image_pipeline, for depth to pointcloud conversion using pinhole camera model. Alternatively, Open3D can be used, but I couldn't find a way to do the conversion and keeping the data structure organized.

So, this implementation does the same thing as the convert function implemented here. Just copy the function in depth_to_pointcloud.py and use it. Features:

• Handles 16 bit unsigned integer depth input.
• Handles 16/32/64 bit floating point depth input.
• Only numpy dependency.