CloudCompare forum

1) I'm trying to register/align clouds from archival imagery to recent lidar-generated. The surfaces are large glaciated mountain faces, and or rock faces where significant rockfall has occurred. It is possible in either case, for me to identify stable areas for point picking. However my question is if there if the use of fine registration will degrade the alignment because it is trying to register areas that are no longer the same? Is there any advice is terms of a workflow involving masking or cutting out those areas?

2) Very basic question, but essential, does CC treat the "aligned" (moving) cloud as a RIGID body (in other words it will stretch/scale it, but the same in every direction? I am correct to assume that the algorithm is fitting one rigid object to the other as closely as possible? No warping?

Thanks!

1) If you want to use ICP in this case, you'll just have to set a low 'overlap' parameter. So that only the common areas are used. And also, make sure to use the entity with the highest density (and widest coverage ideally) as the reference.

2) Yes, it's a rigid registration process, with an optional isotropic scaling (no stretching)