Person authentication using 3D facial data

In this paper an efficient new algorithm is proposed for person authentication based on the analysis of 3D facial data in a cooperative environment. This geometrical approach was motivated by difficulties encountered when considering frontal face recognition. To make the system fast and robust, only the data information of minimum essential visual surface area (two eyes, eyebrows, nose and mouth) is needed for human being to authenticate a person is used in this algorithm. All the data in this area are sectioned vertically and horizontally as profiles and approximated them to minimize database capacity. Two 3D faces are compared with approximated features. Simulations are conducted with real data using our algorithm.