🔒 Locked

Inverse mapping of a quad

Started by pot Jun 2, 2006 at 7:11 AM 1 replies 600+ views

Original Post

Jun 2, 2006 at 7:11 AM

Hi, I am not going to go into too much detail of my implementation, but basically I am inverse mapping to get textures coordinated to re texture a 'morphed' quad. I am having problems when the quad self intersects or the angles at the corners are very acute e.g.

I am implementing the techniques in this paper Fundamentals of Texture Mapping and Image Warping I am implementing the bilinear mapping on page 14 - 17. Im not sure if this is a special case. Does anyone know how to account for these cases? Thanks Pot

DragonL

Jun 4, 2006 at 11:05 PM

What kind of problems are you having? If I may venture a guess, any problems you see probably depend on your implementation (incorrect rounding comes to mind) rather than the algorithm itself.

pot

Jun 5, 2006 at 6:14 PM

Your right about the rounding thing for the acute angle - which is now sorted.

I was having problems with concave / non simple polygons.

I have decided to treat these as special cases and treat them accordingly, i.e. split the quad into two triangles like opengl seems to do.

Topic Locked

This topic has been locked by a moderator. New replies are not allowed.