CGAL users discussion list

[Qianqian Fang <>]

hi

First, a little bit background: my goal is to create a mesh generator 
for meshing
3D regions from the iso-surfaces of volumetric images, see my post at
http://www.nabble.com/failed-to-read-an-OFF-file-tt19423696.html#a19465212
A matlab/octave based meshing toolbox was put forward in Nov. with
work-flows illustrated as in 
http://iso2mesh.sf.net/cgi-bin/index.cgi?Doc/Workflow

In this work-flow, I used a modified surface mesh simplification code
(https://lists-sop.inria.fr/sympa/arc/cgal-discuss/2008-03/msg00170.html)
However, recently, as I worked on more complicated meshes (surfaces from
porous materials or brain cortex surfaces), the self-intersecting 
elements starts
to show up more frequently from the mesh simplifier and break my workflow.
I am trying to either find a robust self-intersecting element removal 
method, or to
modify cgal mesh-simplification code to prevent edge-collapsing if the 
results will
form self-intersecting elements.

As an example, I attached an OFF surface mesh as the raw surface mesh,
to run the mesh simplification by

 cgalsimp2.mexglx  pre_remesh.off  0.05  output_mesh.off

(source code: 
https://orbit.nmr.mgh.harvard.edu/plugins/scmcvs/cvsweb.php/cgalsimp2/?cvsroot=iso2mesh
binaries: 
https://orbit.nmr.mgh.harvard.edu/plugins/scmcvs/cvsweb.php/iso2mesh/bin/?cvsroot=iso2mesh 
)
you will get a surface mesh with 8 intersecting elements (ID: 
66,76,87,72,405,483,489,424)
the raw mesh and the simplified mesh plots are also attached (the 
intersecting
elements are highlighted by red color).

I am wondering if anyone who has similar experiences can give me
some advices, or share some code, on how to implement this feature
with CGAL?

thank you for reading and your feedback will be greatly appreciated.

Qianqian

Attachment: pre_remesh.off.gz
Description: GNU Zip compressed data