Isotropic remeshing returns faces referencing unexisting vertices

[stla]

Hello,

When I run this code (or rather the analogous code in R), then I get a mesh with 2550 vertices but the face indices of this mesh take the values 0->2561 instead of 0->2549. If I remove the faces referencing a vertex in 0->11 and I subtract 12 to the face indices (so the face indices take values 0->2549), then the mesh I get is correct. Note that 12 is the number of vertices of the original mesh.

I already encountered a similar issue when I did some successive intersections of meshes using the same output mesh: there was some kind of cumulation instead of a replacement. I think the above C++ code is correct, this is an issue with R (or with my R/C++ code). Would you have a guess of what happens with my R/C++ implementation?

[sloriot]

Are you aware of the use of garbage in Surface_mesh? See here

[stla]

Thanks. I knew this function, I use it to remove degenerate faces. But I didn't know I had to use it here. I'm going to try now.

[stla]

Yeah, this works!! Thanks!

[afabri]

@sloriot should we add a precondition that there is no garbage, or is it more complex than that (for example in case the users provides an index map)? Is there a warning/note to add in the documentation (here as well as elsewhere)?

[sloriot]

it is documented that indices are in the range [0 -> num_vertices(tm)[

[stla]

Also needed for CGAL::Polygon_mesh_processing::clip.