CloudCompare forum

It really depends on your data (its density, etc.), and of course of the scale of the 'features' you are monitoring.

In your specific case, I would say a scale of 0.05 and a max depth of 0.5 (i.e. something bigger than the box ;). For the normals, I would compute them before hand (with the 'Normals > Compute' tool).

EDIT: you can get very clean normals by using 0.05 as the 'kernel' for normals computation, and -X as preferred direction.

Daniel,

The issue with M3C2 giving me the "CloudCompare has stopped working" error when selecting the guess parameters button is occurring again. It was working about 10 minutes ago and now I cannot get that option to function for me. I am using a very small data set with only a few hundred points so I do not understand why this keeps happening.

Thank you,
CS

This is still a mystery for me as well... I guess you'll have to input the parameters yourself sadly.