"matlab实现指纹识别算法：预处理、特征提取和匹配"

Abstract Fingerprint recognition technology has become one of the most popular and reliable personal identity authentication technologies due to the universality, uniqueness, and invariance of fingerprints. This paper mainly focuses on three aspects of fingerprint image processing: image preprocessing, feature extraction, and feature matching. Image preprocessing includes four steps: image segmentation, filtering enhancement, binarization, and thinning, which removes redundant parts of the original image, facilitating subsequent recognition processing. Feature extraction mainly involves extracting endpoints and bifurcation points of the thinned fingerprint image. Feature matching is the process of comparing feature points of two fingerprint images to determine if they are from the same finger. This paper provides Matlab programs and processing results for fingerprint image preprocessing, feature extraction, and feature matching. The results demonstrate that the processing results of these algorithms implemented in Matlab are ideal, meeting the feasibility and applicability of recognition. Key words: segmentation, binarization, thinning, feature point extraction, matching, Matlab.