Research Of Fingerprint Recognition Algorithms Based On Detail Minutias

With the development of the computer and information technologies, identity identification technology by computer makes quite great progress. Because of the invariability and uniqueness of the fingerprints, the fingerprint verification and identification have become one of the widest used identity identification technologies. It relates to lots of fields, including pattern recognition, digital image processing, digital signal processing, artificial intelligence and so on. It has great worth in criminal identification, network safety, information facility safety etc, and has not only theoretical significance but also practical value.At present, althought the fingerprint identification technology has many finished products, because of the factors of the noises in fingerprints, skin elastic misalignment as well as many literatures don't go to public for commercial interest, identification rate is not very high, the ideal fingerprint identification technology is still a very difficult research assignment.This thesis does deep research mainly on fingerprint identification technology's essential parts,including fingerprint pretreatment, fingerprint feature extraction, fingerprint classification and match. In fingerprint pretreatment part, this thesis does deep analysis and comparison on pretreatment's every step,and has selected an image pretreatment programme.In the feature point information record part, a new method is proposed which uses the ridge count as the distance between two features, even if the distance changes,the ridge count can'change,thus increases the robustness of the system. In the fingerprint classification part, this thesis proposes a new neural network sorting algorithm which bases on the directional diagram characteristic, and has made a very good classification progress.In the fingerprint match part, this thesis uses the fingerprint detail prototype matching algorithm which has been proposed by FBI,it includes three stages: reference point extraction, minutia set of points adjustment and minutia points match. In judging two points whether match is by calculating the differential value of ridge counts between these two points and the reference points separately which whether in the threshold value scope that we have set, instead of calculating the distances between the two points and the reference point. The experiment indicates that this method has obtained good recognition result.