For a series of scattered points with reasonably well defined voids (eg a large cone with a circular apex, finely meshed at the apex, and loosely elsewhere) I am trying to find the constrained triangulation.
It was pointed out to me in the previous post (below) that the new triangulation tools in matlab are better suited to this situation, so I've been trying to work within that framework (2009b currently). I can triangulate the mesh (with no constraints), and use triscatteredinterp to plot, with a reasonable accuracy (it doesn't do as badly on the interior/edge of the original holes as the old griddata for example).
Can we do better though?
What I can't figure out how to do is take this scattered cloud of points and find the constrained region, then do the mesh. I do NOT have access to the exact boundary points off the bat, it needs to be determined dynamically if possible. The demo examples given are not quite getting me there I'm afraid.