[vtkusers] Powercrust weird behaviour.

[IlinkaGrap]

Hi! Recently I have been playing with this
(http://www.sq3.org.uk/powercrust/) Powercrust port to VTK and spotted this
weird behaviour, that occurs on every single model that you try to
reconstruct.

Let me explain this with the case of Stanford Bunny resolution2 surface
reconstruction.

When I reconstruct the Stanford Bunny 3D model(res2) from the ply file which
has 8171 points I get the mesh with 32049! polygons and 46631! vertices. As
far as I understand what Powercrust does is just joining the points of the
originial cloud in the topologically-correct way.

The questions that bothers me is, why do we have 46k number of vertices
instead of 8k. I have to mention that I checked the cloud for duplicates
sorting the std::vector and using std::unique and there were none.
I would also like to add that I face this behaviour with any model I try to
reconstruct using Powercrust.

I would be very grateful if anyone could provide some information on this
case.

Best regards,
Ilia 



--
View this message in context: http://vtk.1045678.n5.nabble.com/Powercrust-weird-behaviour-tp5737680.html
Sent from the VTK - Users mailing list archive at Nabble.com.