DarthPaul

04-19-2004, 01:32 PM

I'm having problems finding papers and websites regarding the removal of self-intersecting polygons in a mesh. I can detect them, but how do I remove them?

View Full Version : Mesh Problem

DarthPaul

04-19-2004, 01:32 PM

I'm having problems finding papers and websites regarding the removal of self-intersecting polygons in a mesh. I can detect them, but how do I remove them?

Tom Nuydens

04-19-2004, 10:55 PM

I'm not sure I fully understand your problem. You want to remove the polygons (or parts of polygons) that are inside of the mesh? If you can detect them, then what's the problem with removing them?

-- Tom

-- Tom

DarthPaul

04-20-2004, 04:54 AM

Ok, so "remove" was the wrong word for me to use. I want to fix the self-intersection.

martin_marinov

04-20-2004, 05:09 AM

Hello,

If you have a general polygon mesh and you want to fix self-intersecting polygons, then you might simply go through every polygon and perform splitting of the faces into smaller parts which does not self-intersect, but only touch. This will solve only *local* self intersection problems.

Global self-intersections are harder to handle, but using a spatial decomposition structure (e.g. BSP-trees) will help here.

At the end of the day, if you wish to perform mesh reconstruction based on polygonal "soups", then volumetric approaches might be of greater help here.

Anyway, posting a bit more information might help us to answer your question. Try for example to specify what kind of mesh you are processing (i.e., general polygons, triangles), what connectivity information (data structure) you are using, what final mesh you would like to produce (2-manifold, triangular, general polygons, geometry tolerances).

Regards

Martin

If you have a general polygon mesh and you want to fix self-intersecting polygons, then you might simply go through every polygon and perform splitting of the faces into smaller parts which does not self-intersect, but only touch. This will solve only *local* self intersection problems.

Global self-intersections are harder to handle, but using a spatial decomposition structure (e.g. BSP-trees) will help here.

At the end of the day, if you wish to perform mesh reconstruction based on polygonal "soups", then volumetric approaches might be of greater help here.

Anyway, posting a bit more information might help us to answer your question. Try for example to specify what kind of mesh you are processing (i.e., general polygons, triangles), what connectivity information (data structure) you are using, what final mesh you would like to produce (2-manifold, triangular, general polygons, geometry tolerances).

Regards

Martin

DarthPaul

05-04-2004, 11:04 AM

I am continuing to work on this problem and I finally found a paper that describes a solution.

Unfortunately the paper is fairly confusing to me. Here's a link to the paper (http://www-unix.mcs.anl.gov/~freitag/Opt-MS/untangle.ps) .

Unfortunately the paper is fairly confusing to me. Here's a link to the paper (http://www-unix.mcs.anl.gov/~freitag/Opt-MS/untangle.ps) .

brcain

05-06-2004, 09:09 AM

I've used Triangle (by Jonathan Richard Shewchuk) to re-mesh sections of a database. It's very fast and easy to integrate into your code (basically two files). E-mail me for an example.

http://www-2.cs.cmu.edu/~quake/triangle.html

http://www-2.cs.cmu.edu/~quake/triangle.html

Powered by vBulletin® Version 4.2.3 Copyright © 2018 vBulletin Solutions, Inc. All rights reserved.