| Re: 3D triangulation of x,y,z vertices [message #50405] |
Mon, 02 October 2006 23:09  |
Dick Jackson
Messages: 347
Registered: August 1998
|
Senior Member |
|
|
Hi Thomas,
Another terminology confusion is that QHull refers to Delaunay triangulation
while, in the case of 3D vertices, it is actually (deep breath, and...)
tetrahedralization. The result is not the triangles of a surface, but the
tetrahedra of the volume within the convex hull of the points.
Your original post:
> I have a 3D object defined by its surface triangles. The connectivity
> is a bit messed up with some faces normals pointing inward (toward the
> inside of the object). Since I have all the points describing the
> surface, I thougth that it would be easy to re-triangulate the x,y,x
> vertices
I see you want neither these tetrahedra nor the convex hull, as you have the
desired *adjacency* info for the triangles of the desired surface and want
the connectivity list for a good polygon mesh with consistent triangle
facing.
If you know that your triangles describe a good mesh (except for
inconsistent facing), you should be able to do this by ensuring that no
segment (between two vertices) is used in the same direction more than once.
If a maximum of two triangles share a segment (as they should in a valid
mesh), one triangle should traverse the vertices as A->B, and the other as
B->A. I put some time into a solution for this, but it's not quite finished
and I have to put it aside right now. Write me directly if you are
interested in what I've done.
Cheers,
-Dick
"Thomas Launey" <t_launey@brain.riken.jp> wrote in message
news:1159796696.413229.92510@m7g2000cwm.googlegroups.com...
> Thanks dick,
> I still have a problem though because "tr" returns the tetrahedra of
> the convex Hull but what I failed to make clear in my post is that I do
> not want the convex hull, only the Delaunay triangulated vertices.
> According to the Qhull doc, when /Delaunay is set, Qhull 'Performs a
> Delaunay triangulation and returns the vertex indices of the resulting
> polyhedra; otherwise, the convex hull of the data are returned. '
> Its seems that even with the /Delaunay switch set, the tr polyhedra
> still represent the convex hull.
> Any clarification or and alternative to re-triangulate 3D vertices
> would be most welcome :-)
> Thanks,
> Thomas
>
> Dick Jackson wrote:
>> Hi Thomas,
>>
>> I think the terminology used in the docs is a bit confusing. QHull's
>> output
>> "Connectivity list" *sounds* like something you'd want for
>> Tetra_Surface's
>> Connin "Tetrahedral connectivity array"... but it's not. What you want to
>> use is QHull's output "Tr" which is a 4-by-nTetra array of indices for
>> each
>> tetrahedron.
>>
>> =====
>>
>> PRO QHullTetra
>>
>> oldverts = RandomU(seed, 3, 20)
>> Qhull, oldverts, tetrahedra, /delaunay
>> newconn=tetra_surface(oldverts, tetrahedra)
>> oPts = Obj_New('IDLgrPolygon', oldverts, Style=0, Thick=3)
>> oSurf = Obj_New('IDLgrPolygon', oldverts, Polygons=newconn,
>> Color=[255,0,0])
>> XObjView, [oPts, oSurf]
>>
>> END
>>
>> =====
>>
>> To see any points hiding inside, choose menu item View:Drag Quality:Low,
>> then press and drag!
>>
>> In IDL, a "connectivity LIST" is a description of a general polygon mesh:
>> [nPts0, <set of "nPts0" indices>, nPts1, <set of "nPts1" indices>, ...]
>>
>> But, as described in Tetra_Clip's doc (but not in all of the
>> tetra-routines'
>> docs, alas):
>>
>> =====
>> A tetrahedral connectivity array consists of groups of four vertex index
>> values. Each set of four index values specifies four vertices which
>> define a
>> single tetrahedron.
>> =====
>>
>> Hope this helps!
>>
>> Cheers,
>> -Dick
>>
>> --
>> Dick Jackson Software Consulting http://www.d-jackson.com
>> Victoria, BC, Canada +1-250-220-6117 dick@d-jackson.com
>>
>
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
