An efficient biometric based personal authentication system using Finger Knuckle Prints features

Abstract

Among several biometric systems presented in the literature, Finger Knuckle Print (FKP) authentication systems have received a great deal of attention in recent years. The present paper investigates a novel method in order to extract the optimal discriminant features from FKP images. This method use the 1D-Log Gabor filter, the Gabor filter bank and the Linear Discriminant Analysis (LDA). In the first step, the Region of Interest (ROI) of a FKP images is analysis with a 1D Log-Gabor wavelet to extract the preliminary feature which is presented by the real parts of the filtered image. In the second step, a Gabor filter bank is applied on the preliminary feature in order to selection only the discriminative features of FKP image. Finally, in the third step, the LDA technique is used to reduce the dimensionality of this feature and enhance its discriminatory power. Our biometric system is based on Nearest Neighbour classifier which uses the cosine Mahalanobis distance for the matching process. Experimental results showed that the proposed system achieves better results than other state-of-the-art systems.