Automated Fingerprint Identification System, A Number Of Key Algorithms

The research of Automatic Fingerprint Identification System (AFIS) is the research pop point in the field of status identification nowadays. Although theoretical research and application development of AFIS have got a significant progress, accuracy of the algorithm and processing speed still need to be improved. In this paper, the structure and working principle of AFIS is firstly described, then by using the knowledge of Digital Image Processing and Pattern Recognition, the problems of fingerprint image enhancement, feature extraction, fingerprint classification and matching are analyzed and discussed in detail. Contributions are listed as follows:1. In the aspect of fingerprint image enhancement algorithm, an improved Gabor-based fingerprint image enhancement method is proposed, which firstly realizes fingerprint image segmentation by linear support vector machine and extracts fingerprint orientation field and ridge frequency, then improved Gabor filter is used to accomplish fingerprint image enhancement.2. In the aspect of fingerprint feature extraction algorithm, an improved minutiae extraction algorithm based on fingerprint gray-scale image is presented. Based on information fusion, it combines the extraction results of fingerprint thinned image to coarsely eliminate the false, then false minutiae in the foreground edge are removed according to segmentation mask, finally several kinds of post-processing rules are integrated to delete different types of false minutiae in fingerprint gray-scale image.3. In the aspect of fingerprint classification algorithm, firstly, combining fingerprint singular point information and central ridge orientation, a fingerprint classification algorithm based on structure and regulation is proposed, which imports model-based feedback mechanism to improve classification accuracy rate. Secondly, a fingerprint indexing algorithm based on average ridge distance is presented, which performs continuous search in the template database using average ridge distance, meanwhile arbitrary classification precision can be realized by adjusting search radius.4. In the aspect of fingerprint matching algorithm, a two-step fingerprint matching algorithm, which combines local matching and global matching, is adopted. A kind of minutiae local structure based on ridge and neighborhood minutiae information is defined in this algorithm. According to the instance of fingerprint core point extraction, different search scheme about the local matching is chosen to reduce computation complexity. After the first step matching, polar coordinates transformation is implemented based on the selected reference point, then elastic matching algorithm is adopted to determine corresponding relation of minutiae's, finally matching score is computed.