Fingerprint cipher design and matching based on orientation pattern

Abstract

First, the image is normalized to improve the clarity of the fingerprint image, and then the fingerprint image is cut to improve the accuracy of feature extraction. Then the fingerprint image is smoothed to reduce the noise. Based on the obvious directionality of fingerprint image orientation, the trend of striations was analyzed by the slicing method, and based on orientation, the binarization refinement operation was carried out to eliminate the skeleton of the fingerprint. When the detail feature points are determined on the fingerprint skeleton image, the relationship between the feature points and the core points can be compared to determine whether the fingerprint matches. Before matching, the effective feature points are obtained by two methods: edge removal and distance removal. After the coordinates of feature points are obtained, a series of simplified and abstract operations are carried out to convert the coordinates into hexadecimal numbers and arrange them according to certain rules. Finally, the "fingerprint password" of less than 200 bytes is obtained.