POSIT params without camera

asked 2013-04-18 05:09:25 -0500

Rogeeeer gravatar image

updated 2013-04-18 05:09:58 -0500

Hi, It may seems weird to use POSIT without a camera but...

I'm actually working on Active Appearance Models for face recognition in order to develop some AR app. So my goal is to have a mesh on the face of my target and add some funny stuff on it.

AAM Fitting take time to be stable and usefull, and I wanted to make a ralistic demo with the generated shapes and textures by varying parameters.

To use POSIT I have to precise intrinsinc parameters, but what is the meaning of these parameters when I don't have any camera ?! I have a 3D model and my points look like this for example :

image description

edit retag flag offensive close merge delete


If it can help someone, I found (for solvePnP but I think it don't really change) : http://www.morethantechnical.com/2012/10/17/head-pose-estimation-with-opencv-opengl-revisited-w-code/ where the author just set fx and fy with image max(w,h), and optical center to cx = w/2, cy = h/2. It makes sense when you think about opengl mvp matrix...

Rogeeeer gravatar imageRogeeeer ( 2013-04-24 09:36:08 -0500 )edit