Let’s say we have a pcd, and I’m handling it with open3d in python.

the ultimate goal is to get a volume of an object extracted as a pcd, but the problem is the pcd values are all normalised, so the calculated volume is something like 0.60.

By the way, I’m using convex_hull to calculate a volume.

It seems like we need to re-convert that using intrinsic or extrinsic camera parameters. But, I’m not sure how this process could be done step by step.

Any suggestion would help me a lot.