8

I've been using one of the older implicit surface algorithms, due to Bloomenthal, as found here, basically the tetrahedral-based algorithm. This works fairly well, but has a shortcoming. Since it uses a fixed grid, it either wastes polygons or ignores detail, depending on the grid size selected.

So my question is, what are my options for improving this? Are there any freely available (source or good description) implicit surface algorithms which are more adaptive to the curvature of the model? Any option I'm missing?

I have found one paper so far which looks promising, pointers to others would be appreciated.

ergosys
  • 47,835
  • 5
  • 49
  • 70
  • In your first link, which document do you refer to? The link goes to a web page with dozens of documents, not a specific one. –  Oct 14 '10 at 03:51
  • @Mads Elvheim, The paper "An Implicit Surface Polygonizer" describes the algorithm. I started with the source in "Public Domain Polygonizer", the C version, which I refactored to C++. – ergosys Oct 14 '10 at 04:34
  • Probably a useless comment, but I have to ask.Can't you simply use a high resolution grid and then simplify the mesh after the fact using a post-step? I don't see how you can do adaptive marching tetrahedrons without knowing something about the implicit function used. –  Oct 15 '10 at 04:02
  • @Mads, This is what I'll do if there is nothing more efficient. Any polygonization algorithm will do, the second paper uses marching triangles. You may be right in the end, but I'm hoping there's a better way. – ergosys Oct 15 '10 at 05:27

1 Answers1

8

AFAIK the following is one of the most influential papers on the field:

Bruno Rodrigues de Araújo and Joaquim Armando Pires Jorge, "Adaptive polygonization of implicit surfaces", Computers & Graphics, Vol. 29, pp. 686–696 (2005)

The method is much better than the "marching tetrahedra" and twice as fast.

alt text

Dr. belisarius
  • 60,527
  • 15
  • 115
  • 190
  • You do know sites go offline eventually, right? The link is broken. The title of the paper could help to find it somewhere else, but you don't even give a hint to what the paper's about. If you could find it again and update this post I'd greatly appreciate it. – Zoomulator Jan 13 '12 at 11:30
  • 2
    @Zoomulator Give this a try: http://sketch.inesc-id.pt/publications/siacg04_brar.pdf – Sciolist Jan 13 '12 at 12:44
  • 1
    Bel, link added :) @zoom see the link. – abcd Jan 15 '12 at 02:58