Automated three-dimensional face authentication & recognition

The needs of security have been increasing therefore organizations have been searching for more secure authentication methods. Face based method is less intrusive and easy to use. The main problems in 2D face recognition are caused by different poses and/or different ambient illumination conditions. These problems can be solved using the 3D face image. This dissertation presents a fully automated 3D face authentication (verification) and recognition (identification) method. The major contributions of this dissertation are: (a) the method can handle data with different facial expressions including hair, upper body, clothing, etc.; (b) development of a robust feature extraction method; and, (c) development of weighted features for discrimination. The input to the face recognition system is a triangular mesh and it outputs a matching rate against a gallery. This method includes both surface and curve based features that are automatically extracted from a given face data. From the experiment, it is concluded that combining the weighted features including profile curves and surfaces gives better authentication and recognition rate than any single feature or combination of weighted features in any feature class.