基于误差扩散的指纹匹配算法model

[中文模板]基于误差扩散的指纹匹配算法郝瑛谭铁牛王蕴红中国科学院自动化所模式识别国家重点实验室北京100080摘要：匹配模块是指纹身份鉴别系统中的核心模块。在采用细节特征描述指纹的系统中，指纹匹配问题就转化为点模式匹配。关键字：误差扩散、指纹匹配、Hough变换1概述给定两个指纹特征模板（参考模板和输入模板），匹配过程通常给出两个指纹模板相似程度的度量。同时，匹配还要设定一个门限，用来确定两个模板是否是从同一个手指提取的。在点模式匹配的理想情况下2算法2.1特征点的相似度度量在介绍特征的相似性度量之前，首先介绍一下指纹特征的表示。在特征提取过程中，为了去除虚假特征点，我们在细化图像上对每个检测出的端点和分叉点所在的纹路进行跟踪。跟踪的结束条件有两个：跟踪长度达到预先设定的某个值或者遇到另外一个特征点。2.1.1对应点估计对应点估计的目的是要找到一对或者若干对最为可靠的对应点。我们首先根据上一节定义的相似度度量，找出所有匹配上的端点对和分叉点对，然后使用Hough变换的方法找出最可靠的匹配点对作为对应点。下面将对具体的算法进行描述。图5.两幅指纹图像的匹配结果，相似度：0.739表1中列出了相同手指的不同样本以及不同手指匹配值的均值。表1.d’以及相同手指和不同手指匹配值的均值和方差数据库d’均值（相同手指）方差（相同手指）均值（不同手指）方差（不同手指）NIST-246.5688.3214.935.659.76参考文献[1]AnilK.Jain,LinHong,SharathPankanti,RuudBolle,“AnIdentity-AuthenticationSystemUsingFingerprint”,Proc.oftheIEEE,Vol.85,No.9,1997.[2]A.Ranade,A.Rosenfeld,“PointPatternMatchingbyRelaxation”,PatternRecognition,Vol.26,No.2,pp.269-276,1993.[3]D.Skea,I.Barrodale,R.Kuwahara,R.Poeckert,“AControlPointMatchingAlgorithm”,PatternRecognition,Vol.26,No.2,pp.269-276,1993.[4]J.P.P.Starink,E.Backer,“FindingPointCorrespondenceUsingSimulatedAnnealing”,PatternRecognition,Vol.28,No.2,pp.231-240,1995.[5]LiHuaZhang,WenLiXu,“PointPatternMatching”,ChineseJournalofComputerScience,Vol.22,No.7,1999.[6]A.K.Hrechak,J.A.McHugh,“Automatedfingerprintrecognitionusingstructuralmatching”,PatternRecognition,Vol.23,pp.7893-904,1990.[7]XudongJiang,Wei-YunYau,“FingerprintMinutiaeMatchingBasedontheLocalandGlobalStructures”,Proc.of15thICPR,pp.1038–1041,2000.[8]Z.Chen,C.H.Kuo,“aTopology-BasedMatchingAlgorithmforFingerprintAuthentication”,Proc.of25thAnnualIEEEInternationalCarnahanConferenceonSecurityTechnology,pp.84-87,1991.[9]D.K.Isenor,S.G.Zaky,“FingerprintIdentificationUsingGraphMatching”,PatternRecognition,Vol.19,pp.111-112,1986.[10]Anil.K.Jain,SalilPrabhakar,LinHong,SharathPankanti,“Filterbank-BasedFingerprintMatching”,IEEETrans.onImageProcessing,Vol.9,No.5,pp.846-859,2000.[11]Chih-JenLee,Sheng-DeWang,“aGaborFilter-BasedApproachtoFingerprintRecognition”,IEEEWorkshoponSignalProcessingSystems(SiPS99),pp.371–378,1999.[12]AnilJain,ArunRoss,SalilPrabhakar,“FingerprintMatchingUsingMinutiaeandTextureFeatures”,Proc.ofICIP,pp.282-285,2001.[13]ZsoltMiklósKovács-Vajna,“AFingerprintVerificationSystemBasedonTriangularMatchingandDynamicTimeWarping”,IEEETran.OnPAMI,Vol.22,No.11,2000.[14]C.Dorai,N.K.Rathat,R.M.Bolle,“DetectingDynamicBehaviorinCompressedFingerprintVideos:Distortion”,Proc.ComputerVisionandPatternRecognition,Vol.2,pp.320-326,2000.[英文模板]Automatic3DFaceVerificationfromRangeDataGangPanandZhaohuiWuInstituteofComputerSystemEngineeringZJU-MiaxisJointLabofEmbededandBiometricsTechnologyZhejiangUniversisty,Hangzhou310027,P.R.China{gpan,wzh}@cs.zju.edu.cnAbstractInthispaper,wepresentedanautomaticapproachfor3Dfaceverificationfromrangedata.Themethodconsistsofrangedataregistrationand3Dfacecomparison.Therearetwostepsinregistrationprocedure.Thecoarsestepconductsthenormalizationbyexploitingaprioriknowledgeofthehumanfaceandfacialfeatures.Keywords:3Dfaceverification,Rangedata1IntroductionTheautomaticfacerecognitionbasedon2Dimageprocessinghasbeenactivelyresearchedinrecentyears,andmanytechniqueshavebeenpresented.Althoughgreatstrideshavebeenmadeduringthepastthreedecades,thetaskofrobustfacerecognitionisstilldifficult.23DfacedatabaseOurexperimentalresultsusethe3Dfacedatafrom“3D_RMA”databaseinM2VTSproject[5].Therangedataareobtainedbya3Dacquisitionsystembasedonstructuredlight.Figure1:TwosamplemodelsinxyzformfromthemanualDB2.1FacedataregistrationA3Dfacerecognitionsystemgenerallymakesupoftwokeyparts:3Ddataregistrationandcomparison.Theaccuracyoftheregistrationwillgreatlyimpactontheresultoffollowingcomparison.2.1.1ThecoarsenormalizationBeforethefineregistrationstep,thecoarsealignmentisperformed.Weassumesthatthegivenrangedataare3Dfacialmodels.Table1:ComparisonofequalerrorratesBeumier[5]OursautomaticDBsession17.25%6.67%automaticDBsession27.75%6.67%automaticDBsession1-29.0%7.33%manualDBsession14.75%3.24%References[1]G.G.Gordon,“FaceRecognitionBasedonDepthMapsandSurfaceCurvature,”SPIEProceedings,Vol.1570:GeometricMethodsinComputerVision,pp.234-247,1991.[2]G.G.Gordon,“FaceRecognitionBasedonDepthandCurvatureFeatures,”Proc.IEEECVPR’92,pp.808-810,June1992.J.C.Lee,E.Milios,“MatchingRangeImagesofHumanFaces,”Proc.ICCV’90,pp.722-726,1990.