Three-Dimensional Face Recognition System - Face Recognition Based on 3D Surface Information

摘要

The performance of face recognition systems that use 2D images is dependent on conditions such as lighting, pose, and facial expressions. We are developing a face recognition system that uses 3D facial surface information to make the system more robust to these arbitrary conditions. We have developed a surface matching system that correlates points in three dimensions between two 3D facial surfaces of different views. We adopt the Iterative Closest Point (ICP) algorithm for surface matching system. Experimental results show that the methods is capable of determining pose and recognition faces accurately over a wide range of poses and with naturally varying illumination conditions.