指纹识别算法的m上atlab实现.doc

PAGE PAGE 4 指纹识别算法的matlab实现 专业：电子信息工程 姓名：马飞 指导老师：刘文博 摘 要 由于指纹所具有的普遍性，唯一性和不变性，以及指纹识别技术具有很高的可行性和实用性，使之成为目前最流行、也最可靠的个人身份认证技术之一。 本文主要对指纹图像进行三方面处理：图像预处理、特征提取和特征匹配。图像预处理包括四个步骤：图像分割、滤波增强、二值化、细化，对指纹图像进行预处理后，去除了原图像的冗余部分，方便后续的识别处理；特征提取主要是提取指纹图像细化后的端点和分叉点；特征匹配是利用两个指纹的图像进行特征点比较，来确定两幅图像是否来自于同一手指。 本文给出了指纹图像预处理、特征提取、特征匹配的matlab程序及处理结果。该结果证明，用matlab实现的这些算法的处理结果比较理想，满足识别的可行性和应用性。 关键词 分割，二值化，细化，特征点提取，匹配，Matlab Abstract Because of the universality, uniqueness and constantness of a fingerprint, and fingerprint identification technology has very high feasibility and practical applicability, make it to be one of the most popular, and most reliable personal identity authentication technology. This paper focuses on three aspects of the fingerprint image processing：image preprocessing, feature extraction, feature matching. Image preprocessing including four steps: image segmentation, filtering, binary, Refining, after The fingerprint image preprocessing, in addition to the original image of redundancy part, convenient subsequent identification processing; The main feature extraction is extracted from the end of the fingerprint image after thinning and bifurcation point; Feature matching is use two fingerprint image feature point is to determine whether the two images from the same finger. This paper provides the fingerprint image preprocessing, feature extraction, feature matching use matlab and handling results, The results prove that these algorithms had ideal results be used by matlab,, Be satisfied with the recognition and feasibility of the application. Key Words: Segmentation, Binary, Refining, Feature point extracting, Matching, Matlab 目录 TOC \o 1-3 \h \z \u HYPERLINK \l _Toc325897976 第1章 绪论 PAGEREF _Toc325897976 \h 4 HYPERLINK \l _Toc325897977 1.1 指纹识别概述 PAGEREF _Toc325897977 \h 4 HYPERLINK \l _Toc325897978 1.1.1 研究背景及意义 PAGEREF _Toc325897978 \h 4 HYPERLINK \l _Toc325897979 1.1.2 国内外研究状况 PAGEREF _Toc325897979 \h 5 HYPERLINK \l _Toc325897980 1.2指纹识别的原理和方法 PAGEREF