The program contains following parts:
- Converting depth map to point cloud
- Mapping 2D color to attach each point in point cloud
- Generating pose axis in point cloud format
- Combining axis with fusion point cloud
More details in Chinese have been uploaded to my CSDN blog.
For your convenience, I uploaded some samples at file source, three random frames are selected from volunteer No. 12.
To run the program in Jupyter notebook, the requirements.txt is already generated. If you want to run in Python, you just need opencv (optional) and open3d.