CGAL users discussion list

[Sebastien Loriot <>]

The best workaround I could advise is to perturb the input point.

Postprocessing the mesh requires removing degenerate faces first
(PMP::remove_degenerate_faces()) before calling the other function.

Best,

Sebastien.

On 11/25/21 6:58 PM, Bruno Manganelli ( via 
cgal-discuss Mailing List) wrote:
> Hi Sebastien,
>
> Glad you could reproduce the issue. Do you have any suggestions for 
> workarounds for the time being?
> I tried 
> using PMP::experimental::autorefine_and_remove_self_intersections but I 
> get an assertion violation.
>
> Thanks,
> Bruno
>
> On Wed, Nov 24, 2021 at 8:26 AM Sebastien Loriot < 
> <>> wrote:
>
>     Thanks I could reproduce the issue.
>     So it turns out that this function does not use the mesher but a
>     dedicated code.
>
>     As it is based on the dual of the Delaunay triangulation and that your
>     points are on a grid, I suspect that the dedicated code does not handle
>     correctly degenerate cases. To check that, I added some noise to the
>     input points and that removed the self-intersecting faces (mostly
>     because of degenerate faces).
>
>     I opened the following issue to track the patching:
>     https://github.com/CGAL/cgal/issues/6147
>     <https://github.com/CGAL/cgal/issues/6147>
>
>     Best,
>
>     Sebastien.
>
>     On 11/23/21 8:55 PM, Bruno Manganelli (
>     <> via
>     cgal-discuss Mailing List) wrote:
>      > Here is a somewhat minimal example.
>      > Hope it's good enough!
>      > Bruno
>      >
>      >
>      > On Tue, Nov 23, 2021 at 5:55 PM Sebastien Loriot
>     < <>
>      > < <>>> wrote:
>      >
>      >     It should not happen. IIRC, it is the output from the
>     Surface_mesher
>      >     so it should be clean.
>      >     Would you have a minimal example showing the problem?
>      >
>      >     Best,
>      >
>      >     Sebastien.
>      >
>      >     On 11/23/21 6:44 PM, Bruno Manganelli
>     ( <>
>      >     <
>     <>> via
>      >     cgal-discuss Mailing List) wrote:
>      >      > Hi,
>      >      >
>      >      > I am using the Union_of_balls_3 class and
>     the mesh_union_of_balls_3
>      >      > function.
>      >      > However, in certain scenarios, the resulting mesh contains
>      >      > self-intersections (PMP::does_self_intersect(mesh) returns
>     true).
>      >      > Is this expected?
>      >      > I can provide an example if required.
>      >      >
>      >      > Thanks!
>      >      > Bruno
>      >      >
>      >      > --
>      >      > You are currently subscribed to cgal-discuss.
>      >      > To unsubscribe or access the archives, go to
>      >      > https://sympa.inria.fr/sympa/info/cgal-discuss
>     <https://sympa.inria.fr/sympa/info/cgal-discuss>
>      >     <https://sympa.inria.fr/sympa/info/cgal-discuss
>     <https://sympa.inria.fr/sympa/info/cgal-discuss>>
>      >      >
>      >
>      >     --
>      >     You are currently subscribed to cgal-discuss.
>      >     To unsubscribe or access the archives, go to
>      > https://sympa.inria.fr/sympa/info/cgal-discuss
>     <https://sympa.inria.fr/sympa/info/cgal-discuss>
>      >     <https://sympa.inria.fr/sympa/info/cgal-discuss
>     <https://sympa.inria.fr/sympa/info/cgal-discuss>>
>      >
>      >
>      >
>      > --
>      > You are currently subscribed to cgal-discuss.
>      > To unsubscribe or access the archives, go to
>      > https://sympa.inria.fr/sympa/info/cgal-discuss
>     <https://sympa.inria.fr/sympa/info/cgal-discuss>
>      >
>
>     -- 
>     You are currently subscribed to cgal-discuss.
>     To unsubscribe or access the archives, go to
>     https://sympa.inria.fr/sympa/info/cgal-discuss
>     <https://sympa.inria.fr/sympa/info/cgal-discuss>
>
>
>
> --
> You are currently subscribed to cgal-discuss.
> To unsubscribe or access the archives, go to
> https://sympa.inria.fr/sympa/info/cgal-discuss