Applying classifier made in CC to a large .pcd in python

[SLAM_enjoyer]

Hi,

I made a classifier using a subset of my data in CC but I am trying to use it on another similar large .pcd file (76 million points) which is not incredibly efficient on CC (takes a while and crashes every once in a while). I was reading through [https://lidar.univ-rennes.fr/en/3dmasc#p-138][/url] and saw that you can import it using cv2, but how would I go about applying the classifier? You probably need to load the point cloud using open3d and convert it to a NumPy array, but what about the features? After I've labelled all the points, how do I create the new point cloud object with labels? I've looked a bit but haven't found such a guide or existing implementation for 3DMASC so any help would be appreciated.

[paul.leroy]

Quite a large amount of questions to answer, some of them not really related to 3DMASC but rather to point clouds and machine learning in Python. Anyway, I put some documentation here, this should help you:
https://lidar-platform.readthedocs.io/e ... dmasc.html