PAGEPAGE44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，MatlabAbstractBecauseoftheuniversality,uniquenessandconstantnessofafingerprint,andfingerprintidentificationtechnologyhasveryhighfeasibilityandpracticalapplicability,makeittobeoneofthemostpopular,andmostreliablepersonalidentityauthenticationtechnology.Thispaperfocusesonthreeaspectsofthefingerprintimageprocessing：imagepreprocessing,featureextraction,featurematching.Imagepreprocessingincludingfoursteps:imagesegmentation,filtering,binary,Refining,afterThefingerprintimagepreprocessing,inadditiontotheoriginalimageofredundancypart,convenientsubsequentidentificationprocessing;Themainfeatureextractionisextractedfromtheendofthefingerprintimageafterthinningandbifurcationpoint;Featurematchingisusetwofingerprintimagefeaturepointistodeterminewhetherthetwoimagesfromthesamefinger.Thispaperprovidesthefingerprintimagepreprocessing,featureextraction,featurematchingusematlabandhandlingresults,Theresultsprovethatthesealgorithmshadidealresultsbeusedbymatlab,,Besatisfiedwiththerecognitionandfeasibilityoftheapplication.KeyWords:Segmentation,Binary,Refining,Featurepointextracting,Matching,Matlab目录TOC\o"1-3"\h\z\uHYPERLINK\l"_Toc325897976"第1章绪论PAGEREF_Toc325897976\h4HYPERLINK\l"_Toc325897977"1.1指纹识别概述PAGEREF_Toc325897977\h4HYPERLINK\l"_Toc325897978"1.1.1研究背景及意义PAGEREF_Toc325897978\h4HYPERLINK\l"_Toc325897979"1.1.2国内外研究状况PAGEREF_Toc325897979\h5HYPERLINK\l"_Toc325897980"1.2指纹识别的原理和方法PAGEREF_Toc325897980\h5HYPERLINK\l"_Toc325897981"1.2.1指纹的基本知识PAGEREF_Toc325897981\h5HYPERLINK\l"_Toc325897982"1.2.2指纹识别的原理及应用PAGEREF_Toc325897982