Biometric Images Processing and Recognition

Abstract

Palm print recognition is physiological biometric method used for recognition of per- son. It consists of various features like texture, statistical, geometry features. Which are used separately or by the combination of two or more features for quick and correct recognition purpose. This technology is beneficial in various fields such as security, access control to buildings, secure payment applications. In this thesis, presents the proposed biometric system, where the deep learning AlexNet is applied on palm-print recognition biometric system. Thus, we based on 2D-DWT as image fusion method and explain their structures. For more comprehen- sive evaluation, we changed parameters and variable of our neural network to selected best combination. Based on the results, excellent accuracy rates war achieved, that provide an overview of the performance of the proposed recognition system.