Automatic 3D face correspondence based on feature extraction in 2D space

This present study proposes to solve 3D faces correspondence in 2D space. Texture maps of 3D models are generated at first by unwrapping 3D faces to 2D space. Following this, we build planar templates based on the 2D average shape computed by a group of annotated texture map. In this paper, landmarks on the unwrapped texture images are located automatically and the final correspondence is built according to the templates. Experimental results show that the presented method is stable and can achieve good matching accuracy.