针对3维人脸识别问题，提出一种由粗到细的两步识别方法。首先结合几何约束与曲率信息定位特征点，根据特征点确定人脸对称面，提取人脸侧面轮廓线。利用轮廓线匹配作为排除算法，在识别初期迅速排除库集中不相似人脸以提高识别效率，剩余库集人脸采用一种具有表情鲁棒性的、基于区域的匹配方法进行识别，该方法自动切割人脸中受表情影响较小的刚性区域，并采用改进的迭代最近点算法对刚性区域进行匹配，为达到更好的识别精度，将各刚性区域的匹配结果采用加法规则融合。在3D_RMA人脸数据库的实验结果表明，该方法具有较好的实时性和鲁棒性。

A two-step matching method for 3D face recognition is proposed. Feature points are detected based on curvature and geometric constraint. Then the symmetrical plane of 3D face is determined based on the feature points, and the profile is determined by the obtained symmetrical plane. The profile is used to form a rejection classifier, which quickly eliminates a large number of candidate faces at an early stage for an efficient recognition in case of large galleries. The remaining faces are then verified using a novel region-based matching approach, which is robust to facial expressions. This approach automatically segments the rigid regions, which are relatively less sensitive to expressions and it matches them separately using a modified iterative closest point (ICP) algorithm. The results of all the matching engines are fused using a sum rule to achieve higher accuracy. Our simulation experiment on 3D_RMA database demonstrates that the proposed method is simple, efficient and robust.