Ask Your Question
0

Direct Occlusion Culling using OpenCV?

asked 2013-05-27 10:02:19 -0600

amir gravatar image

Hi,

I'm new to Computer Graphics and OpenCV. I want to know if OpenCV fits for a direct efficient Occlusion Culling using a GPU. No Rendering is required! Formally:

Input: - list of triangles, representing an 3D object surface. - 3D view point Output: subset of the list which is visible from the given point.

The model is big (about 10-20M triangles) and and need to make it for about 100K different points (possibly in parallel).

If OpenCV fits - where should I start? are there any code samples? I no - any other suggestions?

Thank you!

Cheers, amir

edit retag flag offensive close merge delete

Comments

Sorry, it got wrong, I set the formal requirements again clearly: Input 1: list of triangles, representing an 3D object surface. Input 2: 3D view point. Output: subset of the list which is visible from the given point.

Thanks in advance amir

amir gravatar imageamir ( 2013-05-27 10:07:37 -0600 )edit

2 answers

Sort by ยป oldest newest most voted
0

answered 2013-05-28 08:00:48 -0600

xsmet gravatar image

Short answer: no, OpenCV won't help you here.

Perhaps more helpful: have you looked at a Z-buffer algorithm (a.k.a. visible-surface algorithm)? This is a well documented method of calculating which polygons are visible from a given viewpoint, which sounds like what you need.

There are OpenGL implementations available probably, if you want fast GPU processing.

edit flag offensive delete link more
0

answered 2013-05-27 10:12:04 -0600

berak gravatar image

hi amir,

i don't think, opencv can help you here. it's basically a 2d computer vision library. none of the things you require is built in.

edit flag offensive delete link more

Comments

i'd go and ask again on irc://irc.freenode.net ##opengl ?

berak gravatar imageberak ( 2013-05-27 10:37:03 -0600 )edit

Question Tools

Stats

Asked: 2013-05-27 10:02:19 -0600

Seen: 584 times

Last updated: May 28 '13