Study On Automatic Fingerprint Verification System Based On Point Pattern

Due to the uniqueness and invariability of fingerprint, the automated identification based on fingerprint is becoming an attractive alternative to the traditional methods of identification. The thesis's main point is to study the automated verification system. Based on plenty of literatures, papers and technical reports on fingerprint preprocessing and matching, this dissertation provides AFVS based on point pattern, including two main aspects which are fingerprint preprocessing and matching, respectively. Through experiments, this measure is proved rational and practical. Above all, a new, fast quality assessing algorithm is presented by analyzing direction of fingerprint. Secondly, we synthesize a few measurements to divide the image into unrecoverable regions and recoverable regions. During the segmentation, the paper improves the algorithm of direction. Thirdly, a new binarization method of fingerprint images based on the orientation and the dynamic threshold is proposed, which has the excellent capability of noise resistance. This method makes fully use of the orientation and the characteristics of grayscale change, gets the binarized images from the primitive fingerprint image directly, instead of the series of processing such as smoothing, enhancement and binarization. In the thinning, an improved thinning algorithm is proposed here in order to overcome the shortages of the existing thinning algorithms, a new templates can significantly reduce the scanning iterations and speed up the thinning process while keeping the same thinning results. During the matching, the detail features are extracted using template firstly. There are so many false minutiaes, we must get out them. Then, this paper lodges improved 5-neighbor model. This model use some strong spatial attributes against disturb. There is a good performance in anti-rotation, anti-scale, anti-shift.