The Research Of Fingerprint Feature Extraction And Recognition For Identity Management

Fingerprint is unique and life-long invariance, which makes fingerprint recognition become one of the most reliable, most popular identification. This thesis introduces and analyzes the process of fingerprint identification technology, and a fingerprint identification system. The system includes the following processes: fingerprint image preprocessing, feature extraction and storage, fingerprint recognition. The fingerprint image preprocessing is based on Gabor filterenhancement algorithms and Rosenfeld parallel thinning algorithm for image enhancement and thinning; and then by the means of analysis of the preprocessing results, feature points of images are extracted and saved in a vector; On the basis of feature points, matching algorithms are implemented respectively based on relative distance, quadrant quad tree and image mapping to achieve fingerprint image recognition and identification.