I''m an university student and working on a honour project of human face recognition and 3D modelling. Since I need to make use of some input photos and create a 3D model of that person. I''ve saw some methods on web say that it can done by loaded a generic sample face model first, then reshape the model by the facial points. However, I know how to loaded the 3D model but didn''t know how to work on the reshape of read in model, can anyone help me? Or if any other ways to do it, please tell me.

Thats tricky. For problems like this you need artificial intelligence algoithms - check out www.ai-junkie.com - you probably want a neural network - thats what they currently use to match up two faces. But adapting it for your purpose sounds tricky. In order to do it right you may need to supervise the nets training. Theres a book on it i think - its on the site but you''ll have to but it . You may be able to pull it off with recursion - create a recursive algorithm to divide the face up into flat surfaces depending on the color (similar colors that is).