I am proving the intersection of two tetrahedron using the hemisphere method. (from the book "Game Physics" ) My doubt is : Does it care the direction assigned to each edge when creating the arcs set for each node? (arc between the normal that are shared by the edge). Another question. Which is the easiest (I am beginner) method for intersection test of convex polyhedra that allows me get the contact set ? (Excuse my English. )

Cibressus: I guess you didn't notice that luzop specifically said "Excuse my English." I agree with your point, but it's really a pretty subtle distinction, and not one I'd yell at someone over, particularly not a non-native speaker. (Your English, by the way, luzop, is really very good. I didn't realize you weren't a native speaker until you preemptively apologized for not being one!)

Anyway, to start to answer your second question:

One simple way to test for intersections between convex polyhedra is to make sure that no vertex of either polyhedron is inside the other polyhedron. To test if a given point is inside a given polyhedron, just test that it's on the "in" side of all of the planes that compose that polyhedron. That doesn't get you the contact set though; it's just a true/false intersection test. Someone else may be more helpful here...

I did find an interesting paper on the seperating-axis technique rather than the hemisphere one, though, which may be useful, here.

[edited by - TerranFury on March 22, 2004 11:57:09 AM]

Thanks I appreciate it.

Exactly I was implementing this method.


Thanks again.

Cibressus: