Research Of Palmprint Identification Method Based On Fusion Features And Neural Network

Today, with the rapid development of information and Internet all over the word, information security, especially, personal identification technology becomes extraordinarily important, and must have widely applied perspective. Hereinto, because of its unique features of stability, uniqueness and convenience, the biological feature identification technology has been applied more and more broadly.Biometrics combines the information technology with biology technology, which uses human inherent biological characteristics such as palm-print, iris, and face, and behavioral characteristics such as gait, signature and speech to confirm personal identity for replacing or strengthening the traditional personal identification approaches. Biometric technology is an up-and-coming biology technology and has been applied to feasible systems in the market. It regards its unique, reliable and stable physiological characteristics as the evidence of individual identity and applies the powerful computed ability of computer and technique of network to image processing and pattern recognition, then automated verify the individual identity.Having thoroughly researched the existing palmprint identification technology, in this paper, we give a novel algorithm of key points location, and propose a hierarchical multi-feature scheme to facilitate coarse-to-fine matching for efficient and effective palmprint recognition. In our approach, two-level multimodal features are defined: based on distance of k-means (level-1 feature) and weight-based self-adaptive texture feature based on Zernike moment and ICA through fusion features (level-2 feature). For different features, we adopt two kinds of different identification algorithm, and then combine them into one recognition system effectively. Finally, the experimental results demonstrate the feasibility and efficiency of the proposed system.The main contributions of this paper conclude: (1) Carried on thorough research and compared for the existing palmprint recognition technology. (2) Give a novel algorithm of palmprint key points location and ROI extraction. (3) Both based on distance of k-means and weight-based self-adaptive texture feature based on Zernike moment and ICA through fusion features of BP neural network are as the algorithms of palmprint feature identification and the experimental results show effectiveness and accuracy of the proposed system.