[Insight-users] itk::SimplexMesh and flipped Normals

[Alexandre GOUAILLARD]

dear sebastian,

did you explicitly force consistency of the normal in vtk using
vtkPolyDataNormals::AutoOrientNormalsOn?
if yes, can you check using an itk:QuadEdgeMesh instead of a mesh.
This class will enforce consistent orientation for 2-manifolds. If you
try to pass it a non oriented surface, it will drop the non
consistently oriented cells, and you should be able to see "holes" in
your mesh then.

if both are used and you still have your problem, send me the mesh and
I ll take a look at it.

regards,

alex

On Mon, May 21, 2012 at 11:51 PM, Kiunke, Sebastian
<Sebastian.Kiunke at medizin.uni-leipzig.de> wrote:
>
> Hello,
>
> im currently working on a simplex mesh deformable model.
> My Problem is, that the normals calculated with the "ComputeNormal" function and also if i calculate them in a standard fashion some normals are flipped.
> Thats a problem because the internal forces tend to destruct the mesh and the lighting is not correct.
>
> To get the SimplexMesh i use vtkMarchingCubes to extract a vtkPolyData from a binary image.
> I convert the vtkPolyData to itk::Mesh and then to an itk::SimplexMesh via the known itk classes.
>
> Weird is that the lighting as vtkPolyData is correct.
>
>
> In the itk documentation of itk::TriangleMeshToSimplexMeshFilter is written that "Finally the neighbors of the points are reordered counter-clockwise for geometry computation.".
> If i do the computation of the normal in a consistent way, shouldn't be correct, and ALL normals point in- or outwards?
>
> I would be very very thankfull for help.
>
>
> Greetings,
> Sebastian
> _____________________________________
> Powered by www.kitware.com
>
> Visit other Kitware open-source projects at
> http://www.kitware.com/opensource/opensource.html
>
> Kitware offers ITK Training Courses, for more information visit:
> http://www.kitware.com/products/protraining.php
>
> Please keep messages on-topic and check the ITK FAQ at:
> http://www.itk.org/Wiki/ITK_FAQ
>
> Follow this link to subscribe/unsubscribe:
> http://www.itk.org/mailman/listinfo/insight-users