3D Face Recognition Based On Voronoi Region Integral

Biometric identification is one of the most important characteristics to show machine intelligence and has obtained a wide interest not only in the laboratory but also in civilian application. of all the biometric features, the face is so common and reachable that face recognition remains one of the most active research issues in pattern recognition and image processing. In the past decades, most researches focus on feature extraction from 2D intensity or color images. The recognition accuracy is sensitive to the lighting conditions, expressions, viewing position or variety of subordinates such as hair, glasses. So far, it is very difficult to develop a robust automatic face recognition system.The 3D geometric features can provide more information for recognition than 2D images and some measured features is invariant of transformation, rotation and scale .In addition, 3D capture by some equipments is immune to illumination variation. The 3D information has potential to improve the performance of the system.In this paper, we have proposed a new scheme for face recognition based on Voronoi region integral. In this scheme, the scattered 3D point cloud is first represented with a regular triangulated mesh using generic mesh simplification. Then the 3D geometric features based on Voronoi region integral are extracted . finally, we use the Mahalanobis distance as asimilarity to accomplish the recognition between people after the dimensionality of the geometric feature vector is reduced using principal component analysis. Experimental results on LS_SDU, a 3D face database built by the laboratory of image processing of Shandong University, demonstrate that the proposed algorithm has approving performance.The followings are my main work:(1)In this paper, we establish the first 3D Face Database in China and the proposed algorithm is tested with this database.(2) In this paper, we use the generic head mesh conforming to simplify the original scattered 3D point cloud . The simplification not only conforms all the face meshes, but also reduces the computation and enhances the algorithm' s speed greatly.(3) In this paper, we extract a new 3D global geometric feature based on Voronoi region integral. This feature not only is independent of reference frame, but also can resist voice well. We prove by experiments that this scheme for face recognition using 3D global geometric feature can achieve high recognition rate.(4)In this paper, we use the Mahalanobis distance as the similarity measure and obtain the higher recognition rate than using Euclid distance.