CGAL users discussion list

["Laurent Rineau (GeometryFactory)" <>]

On mercredi 10 août 2011 09:12:55 kyewong wrote:
> Hi, I'm gonna use CGAL::make_surface_mesh to mesh an implicit surface,
> extract the isosurface and convert it to Polyhedron_3 using
> CGAL::output_surface_facets_to_polyhedron.
> I noticed that the algorithm for extracting the isosurface from a given
> implicit surface is the paper <Provably good sampling and meshing of
> surfaces>.
> However, I'm just wondering how about the result and stability of this
> algorithm, comparing to the famous marching cube method used by Jules
> Bloomenthal ? Is it fit for any type of implicit function?
> Thanks!

This algorithm of CGAL::make_surface_mesh is designed for smooth implicit 
surfaces. If your implicit surface is not smooth, then the sharp features of 
the surface will not be meshed correctly. Worst: if your implicit surface is 
not manifold and you activate the manifold criterion, then the algorithm may 
loop and eventually crash.

What is more, if the implicit surface has several connected components, the 
result of the meshing is not guaranteed to mesh all connected components of 
the surface. The initial step should be carreful to add a few points to every 
connected components.
 
-- 
Laurent Rineau, PhD
R&D Engineer at GeometryFactory           http://www.geometryfactory.com/
Release Manager of the CGAL Project       http://www.cgal.org/