Three Dimensional Palmprint Recognition Using Linear Discriminant Analysis Method

As a significant biometric technique, 3D palmprint authentication is better than 2D palmprint authentication in several aspects. Previous work on 3D palmprint recognition has concentrated on two aspects: (1) extracting the texture and line features using the binary image of 3D palmprint; (2) extracting the orientation features using the Gabor filter and competitive code. In this paper we extract, for the first time, the 3D palmprint features using the appearance-based linear discriminant analysis (LDA) method. The appearance-based LDA method can extract the global algebraic features of the biometrics. These features have been proven to have strong discriminability. We also investigated the relationship between the recognition accuracy and the resolution of the 3D palmprint image. The experimental results show that the 3D palmprint images with resolution 16×16 and 32×32 are better for 3D palmprint recognition. The experiment results also show the feasibility of our method.