3D face recognition based on intrinsic features

Abstract

Changes in human expression cause non-rigid deformation of face models, this is a great challenge for 3D Face Recognition. To tackle this problem, there has been lots of excellent research work in recent years. In this paper, we propose a face recognition algorithm based on intrinsic features. Firstly face models are preprocessed and adjusted to standard pose for extracting nose tip, then we compute several geodesic stripes based on detected nose tip, make sampling in each stripe, and extract isometric-invariant features on each feature point. Because facial expression makes different levels of impact on different parts of face surface, we use SVM to train the matching results between stripes, getting optimal weight for each stripe. Finally, similarities are computed by weighed sum of different stripes matching results. Our experiments use the Gavab Database and the results are better than other 3D face recognition algorithms such as MDS method, showing effectiveness of our method. © Springer International Publishing 2013.