CGAL users discussion list

["Sebastien Loriot (GeometryFactory)" <>]

Michel Audette wrote:
> Dear CGAL developers,
>
> I'm trying to use the edge_collapse() surface mesh decimation method in 
> conjunction with the output, in off format, of GNU Triangulated Surface 
> Library, to compare the results of VTK's Marching Cubes with GTS' Dual 
> Marching Tetrahedrons. The GTS surface is a little denser than the VTK 
> Marching Cubes result: 18M polygons vs 7M polygons, and currently does 
> not seem to produce an output..
>
> michel@michel-desktop:~/simProject/NeuroSurgSim/SimplexDecimationBuild/bin$ 
> ./OffMarchingCubesCGALDecimate brainDualMarchingTet.off 0.002 
> brainDualMarchingTet_0_002
> Reading polyhedron ...
>     .... done.
>
> Finished...
> 0 edges removed.
> 0 final edges.
The 0 final edges is strange. Maybe you should check that the input has 
been correctly read.


> Is there a verbose way of using the function that can provide hints? 

Before any CGAL include file add
#define CGAL_SURFACE_SIMPLIFICATION_ENABLE_TRACE 4
void Surface_simplification_external_trace( std::string s )
{
  static std::ofstream out("log.txt");
  out << s << std::endl ;
}

and in the main before calling simplification code:
internal::cgal_enable_ecms_trace = true ;

Also you need to use 
CGAL::Polyhedron_3<Kernel,CGAL::Polyhedron_items_with_id_3>
as polyhedron type.

you will then have a log.txt. I will provide no support on
information in this log file. You can play with the value of
CGAL_SURFACE_SIMPLIFICATION_ENABLE_TRACE to print more or less
information (the greater the value is, more information you will get)

S.

> Does 18M faces seem like a lot for edge_collapse to process?
>
> Best wishes,
>
> Michel
>
>
> --
> Michel Audette, Ph.D.
> R & D Engineer,
> Kitware Inc.,
> Chapel Hill, N.C.