Palmprint Recognition System by Using Contourlets Transform and Artificial Neural Network

Abstract

Palmprint recognition is a promising biometric system for forensic and commercial applications. Palmprint recognition has been investigated over ten years. Many different problems related to palmprint have been addressed. The proposed work provides a novel technique from the combination between multiscale transforms: 2D discrete wavelet, Ridgelet, Curvelet, and Contourlet for features extraction phase, 2D principal component analysis (2D PCA) and 2D linear discriminant analysis (2D LDA) are used for dimensionality reduction. Feed forward back-propagation neural network used for recognition phase. The algorithms have been tested by using PolyU hyperspectral palmprint database. Finally, a comparative analysis of the proposed approach has been taken. The recognition accuracy of the proposed approach with some existing approaches are taken. The simulation results show that the proposed approach gives good and comparable results with some other approaches.