Preprocessing On Fingerprint Image And Minutiae Extraction

There is no two fingerprints are exactly the same, this point has been accepted by most people, and due to their uniqueness and persistence, fingerprints are used as main basis of personal identity. Fingerprint identification has been applied for a long time. At present, it has been adopted by the government and becomes an effective identity recognition approach in law. Fingerprint identification is of convenience, high efficiency, security and reliability. It has also been applied in many fields such as financial security, data encryption, and electronical business and will play a more and more important role in our future life necessarily.Fingerprint identification is usually made up of several sections: preprocessing, thinning, post-processing, minutiae extraction, minutiae matching. This paper introduces the current status and correlative techniques about fingerprint identification according to above-mentioned processes at first, discusses the structure and its algorithm in AFIS in detail, analyses its working principle and each process of preprocessing also. Basing on these a set of preprocessing algorithms is redesigned. Enhancement and binarization algorithm are the emphases among them. The filter mask is determined by fingerprint images' orientation, application of the filter shows good ridge separation and continuity, and background noise reduction. Then by dynamic threshold binarization algorithm, the gray images get binarized.At last, applying different algorithms to get rid of all kinds of false features on thinned fingerprint images according to their particular characteristics, namely post-processing, clear and exact fingerprint point-and-figure charts are received. Minutiae extraction algorithm is achieved through the coding of eight-points surrounding which show the type of aim point and the states of its neighbored points, thus it can be seen the aim point is the ending or the bifurcation, and then in term of the grey values of eight-points surrounding minutiae can be extracted.AFIS is realized by visual C++ on windows platform. Experiments have been done and the results show that the algorithm dealing with fingerprint images is robust, it can make the images get enhancement effective and eliminate most fake minutiae.