CloudCompare forum

Hi all,
I'm using the M3C2 algorithm in order to realize a laboratory experiment in which I want to simulate an opening fracture between two rocks. I have got four different aligned point clouds of the two rocks representing four different steps. In every step I have increased the fracture of 1 cm moving only one rock. In the M3C2 plugin I have used the first step as the reference cloud upon which I have compared the others clouds. Is it right to project the M3C2 distances on the reference cloud to evaluate the displacement between the two rocks in two different steps? Then, I analyzed the gaussian distribution of every comparison between two clouds and I have seen that the negative values are distributed on the stable rock and the positive values on the moving rock and in the fracture. In your opinion is it correct to interpretate the positive part of the gaussian distribution as the forced displacement in the second rock and in the fracture? Thank you very much indeed for your help.

Best regards
Sara

Yes, with M3C2 it's valid to project the distances on any cloud you want (assuming that the correspondance between both clouds is correct - for this you'll have to carefully chose the way the normals are computed).

And for the other part of your question, can you post a snapshot of a cloud with distances? (at least for one epoch).

By the way, how are the clouds "registered"? Did the coordinate system / sensor remained fixed all the time?

Hi Daniel,
the clouds are scaled and aligned using the cloud of the step zero as reference. As you asked me I'm going to post you a snapshot for the first epoch that consist of a displacement of 1 cm. The M3C2 algorithm should register this displacement on the visible rock in the shot, which is the moved rock. The other one is stable in the experiment. I have reduced the gaussian distribution considering the interval from 0.0018 m to 0.01 m; the first one value is the final RMS derived from the two aligned clouds, the second value is the manual displacement induced on the rock. Do you think that this method to evaluate the changing distance between the two clouds and the opening fracture could be correct?
Thank you very much indeed for your help.
Sara

Could you zoom in so that we can see the cloud in a better way? Your normals seems a bit crooked (especially the black ones).

Why and how exactly are you registering the clouds? Because depending on how you do it it can invalidate the whole process.

Maybe it would be better to communicate by email. I would like to see the two clouds (before and after) and try to understand how the displacement exactly occurs, etc.

Hi Daniel,
thank you very much indeed for your help. I'm going to contact you in private attachig some shots.

Sara