Research On Preprocessing Algorithm For Fingerprint Verification System Based On ARM7

Two fingerprints cannot be exactly the same, which has been widely acknowledged, and due to their uniqueness and persistence, fingerprints are used as main basis of personal identity. In the past two decades, automatic fingerprint verification technology has developed rapidly and been applied in various fields. In this paper, Fingerprint patrol management system based on ARM7(S3C44B0X) is introduced, which needs high requirement on time complexity of fingerprint verification algorithm. Based on many literatures studying, image preprocessing with highest time complexity in automatic fingerprint verification system is analyzed in detail. The scaling-processing program to reduce preprocessing time is put forward.Image preprocessing is usually consisted of following steps: quality assessment, mean filtering, gray normalization, gray equalization, the oriental filtering, binarization as well as thinning. In this paper, current development situation and relative techniques about fingerprint verification are introduced at first. Secondly, structure and algorithm of automatic fingerprint verification system are discussed in detail, including working principle and each process of preprocessing. Enhancement and binarization algorithm are the emphases among them. According to fingerprint feature, orientation map is calculated, and image is denoised by orientation filters. Results of filtering show good ridge separation, continuity and noise reduction. Then by means of dynamic threshold binarization algorithm, the gray image is binarized. Based on these, a set of scaling-processing program is redesigned. Time and matching performance parameters are shown. Final experiments prove that the program has greatly reduced preprocessing time without changing the matching performance.The system is realized by visual C++ on Windows platform, and transplanted to ads 1.2(ARM Development Suite). Results of experiments show that the algorithm dealing with fingerprint images is robust. It can make the images enhanced effectively.