CGAL users discussion list

[Mael Rouxel-Labbé <>]

Hello,

Note that smooth_mesh() is intended to improve the quality of the elements on the discretization of a surface, not to change the shape of the surface itself (although that some change will happen due to re-projections on the initial surface happening within smooth_mesh()).

PMP::duplicate_non_manifold_vertices() indeed creates new vertices but does not change the geometry. There is some on-going work to repair on manifold vertices / edges, but it is not ready to be used just yet. It seems that the experimental self-intersection function that Maxime pointed out in a previous message would best fit your needs for now as it will also handle non-manifold vertices (after duplication).

Best,
Mael

On 15/07/2020 20:31, Yaoyu Hu ( via cgal-discuss Mailing List) wrote:

Thank you Maxime and Mael! The Euler::remove_face() finally works. I was so silly when I first attempted to use Euler::remove_face(), where I remove more elements than I really wanted. Now the downstream smoothing works fine. 

I also tried Mael's suggestion and apply duplicate_non_manifold_vertices(). As long as I understand, this function will duplicate the non-manifold vertices in the same spatial positions. Say, a single non-manifold vertex becomes multiple manifold vertices, am I right? If this is the case, how can I smooth out the local regions around the original non-manifold vertices, to make the region more smooth? It seems to me that the facets associated with the newly duplicated vertices and the nearby facets are still arranged in a pattern much more complex than a plain surface.

In the sample surface mesh I enclosed in the previous email, there are some wired facets locate around the surfaces near the top. These facets consist of most of the original self-intersections and non-manifold vertices. After I remove the self-intersecting facets and duplicate the non-manifold vertices, but the remaining facets are still in wired positions. How can I further process the mesh so that these bad regions become more like a plain surface?

Thank you for the help!

        -- 
        You are currently subscribed to cgal-discuss.
        To unsubscribe or access the archives, go to
        https://sympa.inria.fr/sympa/info/cgal-discuss