Closed dennisushi closed 7 months ago
Hi, In our experiments, we do calibration to get the camera extrinsics to transform the point cloud into world frames. I think it is ok if you just use depth + intrinsics and do not calibrate (from our experiments it could work also but maybe some performance drop generally)
Following the Converstion script, is it strictly necessary to transform the pointcloud to world frame (i.e. we need to do external eye to hand calibration) or not?