Threshold too high
Pippo809 opened this issue · comments
I am trying to compute the convex decomposition of a mechanism but the threshold parameter is too high and does not allow to obtain a usable result.
What can I do to obtain a better result?
Have you ever tried a lower concavity threshold by using -t option?
Yes, those are the results with the lowest setting possible (-t 0.01) I was wondering why there wasn't an option to go lower
Would you mind sharing your input files?
Hi the pre-processing will remesh non-manifold mesh and may lose some details. So the problem may not be caused by the threshold. If your mesh is manifold, you can use -np to turn off the pre-process; if not, you can try to increase the remeshing resolution by -pr.
As @SarahWeiii commented, it may be mostly problem of the non-manifold preprocessing. In some applications, I tried using CoACD without pre-processing for general meshes (manifold and non-manifold) and it works in most cases. In the worst of the cases, it diverges a little bit, but it converges eventually for every mesh.
If not using pre-processing for non-manifold meshes, there may be trouble in the triangulation of newly generated parts during cutting, which can lead to redundant convex hulls. But in general, it should not affect the approximation of the original shape.