ICP after georeference

[yangmaximum]

Hi Daniel,

I am a master student in geotechnical engineering. I am currently analyzing point cloud models. We have four stages of diaphragm wall point cloud models. We hope to use the control points for georeference registration.However, due to the accuracy of these control points are not good(about 3cm), and for the local soil properties, the expected displacement of the diaphragm wall is only about 15mm, which means that the analysis results will be meaningless.

Therefore, I plan to register the point cloud by icp after it has been georeferenced, and then use cloudcompare for M3C2 analysis, but at this time I have a question:

Since there will be displacements between the point clouds at each stage due to the deformation of the diaphragm wall, if we perform ICP again, will the "deformation of the diaphragm wall" that we are interested in completely disappear?

I don't know if I'm expressing it well, I look forward to seeing your professional reply, thank you.[/img]

[daniel]

Are there some parts that didn't move? (or at least you are confident they didn't)

In this case it's possible to isolate these parts (with th scissors tool), then use them to register, and eventually copy the resulting transformation matrix in the Console (CTRL+C) and apply it on the original cloud with 'Edit > Apply Transformation'.