Fingerprint Recognition Based On The Similar Principle Of Triangle

Due to the uniqueness and invariability, the fingerprint(FP) provides an optimal foundation for Automatic Personal Identification System (APIS). And fingerprints identification is becoming one of most attractive, convenient and reliable technology of identification. Although there exists many product for fingerprints identification, the design of ideal automated fingerprint identification system (AFIS) is still a difficult research subject by reason of much special noise in fingerprint images.This paper focuses on one of the most important techniques in automatic fingerprint identification system: Fingerprint Matching. Identification of the two FP which is in different size, different angle and position, some solution will unify their size, angle and position first. It make the algorithm very complicated and lower the accuracy of the identification.On the basis of the analysis on traditional identification, a new fingerprint identification solution based on the similar principle of triangle is approached. An AFIS is realized by using Visual C++.NET on computer which is made up of the fingerprint image preprocess and character identification. There are five processes in the preprocess part, which are image normalizing, image filter, orientation estimation, binarization and thinning. There are two parts in the stage of minutiae extraction ,one is bifurcation extraction and the other one is post-process of the thinned FP image. For testing the efficiency of the matching algorithm, many pertinency experiments are taken. These experiments include the matching of the FP images which come from different fingers, the matching of the different FP images which come from a same finger; the matching of FP images which is not in same size and the matching of one to many. The results of experiments demonstrate that the method doesn't rely on the core points, and furthermore, it is invariant to rotation and zoom.