Fingerprint Identification System Algorithm

Fingerprint Recognition has been widely used in security area for its uniqueness and reliability. This thesis has discussed the problems of fingerprint image segmentation and enhancement, fingerprint minutiae extraction and verification, fingerprint minutiae matching, using the knowledge of digital image processing, pattern recognition and computational intelligence.The major contents of this thesis are listed in the following:(1) Segment original fingerprint image based on the Canny operator to detect the image's edge, and then enhance fingerprint image using nonlinear fusion filter.(2) Estimate fingerprint image point orient field and block orient field, then binarize the fingerprint image based on its block orient field.(3) Thin the BW fingerprint image using the Hilditch method, and extract the original fingerprint minutiae based on the thinned image.(4) Analyze various noises in fingerprint images and their properties, and generalize distributing regulation of pseudo minutiae. Combining with the information of local ridge direction, special algorithms are designed with respect to various noises to delete pseudo minutiae from original minutiae set.(5) Discuss a fingerprint matching algorithms based on minutiae, correct template data and input data, take bound box of variable size to obtain matching score to identify whether they belong to a same fingerprint. The algorithm is not depended on center area of fingerprint image, and becomes more robust on nonlinear transfigure.