CGAL users discussion list

[SirM2X <>]

I'm facing issues with mesh deformation algorithms in CGAL. I read through 
this document
<http://doc.cgal.org/latest/Surface_mesh_deformation/index.html>   to get a
good understanding of how to use the deformation module.

What I'm trying to do, is deform a mesh by specifying the whole mesh as the
ROI and specify some silhouette points of the mesh as control points, and
deform the mesh by specifying the target position of the control points.

For reference, the image below summarizes what I want to do with the cactus
mesh (cactus.off) that's bundled with CGAL:

<http://cgal-discuss.949826.n4.nabble.com/file/n4662924/cactus-front.png> 

White points are vertices of the cactus's mesh. Green points are the control
vertices on the silhouette of the mesh, viewed from the current position
(these points are extracted  using the method in this paper)
<https://www.cs.rutgers.edu/~decarlo/pubs/sg03.pdf>  . There are
approximately 60 green vertices. Red points are the target position for the
control vertices. Each red point is associated with a green vertex and they
are located nearby each control point.

To illustrate things better, here is a side view of the above figure:

<http://cgal-discuss.949826.n4.nabble.com/file/n4662924/cactus-side.png> 

I've tried variations of the deformation algorithm (spokes and rims, ARAP,
etc) with different parameters, yet all algorithms seem to produce horrible
results. Below is a sample result, produced by a code excerpt that's pasted
at the bottom of this post. The first image is the vertices of the deformed
mesh (again, target position of the control points are signified by red),
and the second image is the output mesh opened in MeshLab:

<http://cgal-discuss.949826.n4.nabble.com/file/n4662924/cactus-deformed.png> 

<http://cgal-discuss.949826.n4.nabble.com/file/n4662924/cactus-mesh.png> 

As you can see, not only the control vertices are positioned weirdly, but
also the output mesh is completely weird looking (looks like the Iron Throne
in the Game of Thrones )

I've tried playing with the parameters (iterations, tolerance, etc) to no
avail. I appreciate any help or insight on what may be going on here...

The code that produced the above mesh is as follows:







--
View this message in context: 
http://cgal-discuss.949826.n4.nabble.com/Mesh-deformation-produces-strange-results-tp4662924.html
Sent from the cgal-discuss mailing list archive at Nabble.com.