Research Of Pressure Fingerprint Identification System Key Technology

Nowadays,with the great progress of science technology and the diversification of information acquisition mode,more and more people becomes having a great demand of the safe and reliable identification products. Owing to fingerprint identification technology has the characteristics of high feasibility and high availability which is not only convenient and practical but also is the most popular and widely used techniques of identification at home and abroad.In recent years, key technologies of the fingerprint recognition have been studied by many research institutions domestic and overseas, and continue to be a study in depth. It aims at improving the automated fingerprint identification system performance continuously to meet the requirements of practical using, especially, getting out of the traditional PC, embedded devices which have independence and flexible, real-time efficient and low power consumption and small size. The core technology of fingerprint identification with independent intellectual property rights will become the high and new technology industries in S&T development project programs. The main contributions of this paper are described as follow:(1) In acquisition of the fingerprint image, this paper uses combination of Japan BMF's BLP100 pressure sensitive fingerprint sensor and control chips BCT100, through the simulation software to determine the best option and design proper PCB. The core processor of Recognition system adopt the TI's DSP5502, and achieve acquisition system initialization and acquisition program designed in CCS.(2) The pre-processing algorithms include normalization, cropping, enhancement, binarization, thinning, etc. Firstly, this paper analysis factors which affected source fingerprint image quality during the process of fingerprint images collected. And then base on the actual requirement of this system, by comparison and analysis of various processing algorithms. The paper uses a number of efficient processing algorithms and improves part of the preprocessing algorithm, including the improvement of the direction and frequency calculation based on the original Gabor filter and analyzing the shortcomings of the OPTA algorithm and its improvement, this paper further improves the OPTA algorithm through the template expansion, etc. (3) According to the actual needs of the application, this system accomplishes the feature extraction and matching algorithms. The most important aspect of this paper is analysis and processing the pseudo minutiae which arises from image thinning and giving the improved method of the false feature eliminating. On the basis of summary and comparison previous studies, fingerprint matching algorithm used in this paper is congruent triangle matching algorithm, which has the features of quick run speed and high precision.(4) This paper completes a set of processing algorithms and achieves the integration of hardware and software. Through the test by Zhejiang Fangyuan Test Goup and Jiaxing FuBusen Biometric Technology Co. Ltd which proves the system is feasible.