Mutli-Modal Face Recognition

Most of the approaches that have been developed for 3-D face recognition are based on 3-D surface matching. I presented an approach for 3-D face recognition based on Ridge lines extracted from range facial images. Only the points around the important facial regions on the face (i.e., the eyes, the nose, and the mouth) were used, and the surface patches on the face were ignored during the matching process. I showed by experiments on the Face Recognition Grand Challenge (FRGC) database, the largest available 3-D face database, that the ridge lines carry the most important information on the face surface for face recognition.