I see.  One thing you might want to try is first getting a triangulation (even 
a bad one) and then using gmsh's surface remeshing algorithm on it.  I've done 
this on one project and the results are quite good. 

Dave

On Aug 23, 2013, at 9:42 AM, Jack Stalnaker <[email protected]> wrote:

> Thanks,
> 
> That's what I figured. I've gone that route already, using PCL to triangulate 
> the surface using their greedy projection algorithm. The problem is that it's 
> quite slow for my data sets. I have tried flattening and triangulating, too. 
> That's fast, but introduces bad triangles. I noticed a lot of geoscientists 
> on this mailing list, and thought someone might have a simpler solution for 
> going from seismically derived points representing a reflector to a mesh 
> surface. 
> 
> --Jack
> 
> 
> 
> On Fri, Aug 23, 2013 at 11:25 AM, David Bernstein <[email protected]> wrote:
> Hi Jack, I'm assuming the points which define the top surface are not already 
> on a simple surface like a plane or a sphere.  I guess your problem is that 
> you have to define a surface which contains the points and let gmsh know what 
> the surface is.  One way to do this is to create a triangulation of the point 
> set and then define lines, line loops, and surfaces from the triangles and 
> then create a surface loop out of all the triangles.  However, I don't know 
> of an algorithm that will create a triangulation of a set of points in 3D in 
> such a way that it makes a sensible surface (I'm not an expert in this area 
> though) so you might have to do the initial triangulation manually.
> 
> Regards,
> Dave
> 
> 
> On Aug 22, 2013, at 12:11 PM, Jack Stalnaker <[email protected]> wrote:
> 
> > (Apologies. I hit send before completing the email before)
> >
> > I have a 3D body defined as follows:
> >
> >      - bottom, front, back, and sides are flat
> >      - top is defined by a set of points in 3D
> >
> > How do I go about creating a 3D unstructured mesh from this body using 
> > Gmsh? The top surface is the problem. It is only defined as points. Do I 
> > need to define lines and line loops from these points in order to create a 
> > mesh? Is there some standard way to do this in Gmsh, like a built in tool?
> >
> > Thanks,
> > Jack
> > _______________________________________________
> > gmsh mailing list
> > [email protected]
> > http://www.geuz.org/mailman/listinfo/gmsh
> 
> 
> _______________________________________________
> gmsh mailing list
> [email protected]
> http://www.geuz.org/mailman/listinfo/gmsh
> 
> _______________________________________________
> gmsh mailing list
> [email protected]
> http://www.geuz.org/mailman/listinfo/gmsh

_______________________________________________
gmsh mailing list
[email protected]
http://www.geuz.org/mailman/listinfo/gmsh

Reply via email to