文献翻译——基于细节分数匹配法的指纹识别.doc
- 文档编号:96633
- 上传时间:2022-10-02
- 格式:DOC
- 页数:23
- 大小:150KB
文献翻译——基于细节分数匹配法的指纹识别.doc
《文献翻译——基于细节分数匹配法的指纹识别.doc》由会员分享,可在线阅读,更多相关《文献翻译——基于细节分数匹配法的指纹识别.doc(23页珍藏版)》请在冰豆网上搜索。
外文文献资料
FINGERPRINTRECOGNITIONUSINGMINUTIASCOREMATCHINGABSTRACT:
ThepopularBiometricusedtoauthenticateapersonisFingerprintwhichisuniqueandpermanentthroughoutaperson’slife.Aminutiamatchingiswidelyusedforfingerprintrecognitionandcanbeclassifiedasridgeendingandridgebifurcation.InthispaperweprojectedFingerprintRecognitionusingMinutiaScoreMatchingmethod(FRMSM).ForFingerprintthinning,theBlockFilterisused,whichscanstheimageattheboundarytopreservesthequalityoftheimageandextracttheminutiaefromthethinnedimage.Thefalsematchingratioisbettercomparedtotheexistingalgorithm.
Key-words:
-FingerprintRecognition,Binarization,BlockFilterMethod,MatchingscoreandMinutia.
1.Introduction
Biometricsystemsoperateonbehavioralandphysiologicalbiometricdatatoidentifyaperson.Thebehavioralbiometricparametersaresignature,gait,speechandkeystroke,theseparameterschangewithageandenvironment.Howeverphysiologicalcharacteristicssuchasface,fingerprint,palmprintandirisremainsunchangedthroughoutthelifetimeofaperson.Thebiometricsystemoperatesasverificationmodeoridentificationmodedependingontherequirementofanapplication.Theverificationmodevalidatesaperson’sidentitybycomparingcapturedbiometricdatawithreadymadetemplate.Theidentificationmoderecognizesaperson’sidentitybyperformingmatchesagainstmultiplefingerprintbiometrictemplates.Fingerprintsarewidelyusedindailylifeformorethan100yearsduetoitsfeasibility,distinctiveness,permanence,accuracy,reliability,andacceptability.Fingerprintisapatternofridges,furrowsand
minutiae,whichareextractedusinginkedimpressiononapaperorsensors.Agoodqualityfingerprintcontains25to80minutiaedependingonsensorresolutionandfingerplacementonthesensor.Thefalseminutiaearethefalseridgebreaksduetoinsufficientamountofinkandcross-connectionsduetooverinking.Itisdifficulttoextractreliablyminutiafrompoorqualityfingerprintimpressionsarisingfromverydryfingersandfingersmutilatedbyscars,scratchesduetoaccidents,injuries.MinutiabasedfingerprintrecognitionconsistsofThinning,Minutiaeextraction,MinutiaematchingandComputingmatchingscore.Motivation:
Themotivationbehindtheworkisgrowingneedtoidentifyapersonforsecurity.Thefingerprintisoneofthepopularbiometricmethodsusedtoauthenticatehumanbeing.TheproposedfingerprintverificationFRMSMprovidesreliableandbetterperformancethantheexistingtechnique.
Contribution:
InthispaperweusedFingerprintRecognitionusingMinutiaScoreMatchingmethodwiththehelpofMATLABcodes.Minutiaeareextractedfromthethinnedimageforbothtemplateandinput
image.Finallyboththeimagesaresubjectedtomatchingprocessandmatchingscoreiscomputed.
Organization:
Thispaperisorganizedintothefollowingsections.SectionIIisandefinitionoftherelatedworkanddescribesModelforfingerprintrecognitionindetail,SectionIIIgivesthealgorithm.InsectionIVperformanceanalysisandresultsarediscussedandfinallyinsectionVgivetheconclusions
2.Relatedwork
G.SambasivaRaoetal.,proposedfingerprintidentificationtechniqueusingagraylevelwatershedmethodtofindouttheridgespresentonafingerprintimagebydirectlyscannedfingerprintsorinked36impression.RobertHastingsdevelopedamethodforenhancingtheridgepatternbyusingaprocessoforienteddiffusionbyadaptationofanisotropicdiffusiontosmooththeimageinthedirectionparalleltotheridgeflow.Theimageintensityvariessmoothlyasonetraversealongtheridgesorvalleysbyremovingmostofthesmallirregularitiesandbreaksbutwiththeidentityoftheindividualridgesandvalleyspreserved.JinweiGu,
etal.,proposedamethodforfingerprintverificationwhichincludesbothminutiaeandmodelbasedorientationfieldisused.Itgivesrobustdiscriminatoryinformationotherthanminutiaepoints.Fingerprintmatchingisdonebycombiningthedecisionsofthematchersbasedontheorientationfieldandminutiae.V.VijayaKumariandN.SuriyanarayananproposedamethodforperformancemeasureoflocaloperatorsinfingerprintbydetectingtheedgesoffingerprintimagesusingfivelocaloperatorsnamelySobel,Roberts,Prewitt,CannyandLoG.Theedgedetectedimageisfurthersegmentedtoextractindividualsegmentsfromtheimage.RajuSonavane,andB.S.Sawantpresentedamethodbyintroducingaspecialdomainfingerprintenhancementmethodwhichdecomposesthefingerprintimageintoasetoffilteredimagesthenorientationfieldisestimated.Aqualitymaskdistinguishestherecoverableandunrecoverablecorruptedregionsintheinputimagearegenerated.Usingtheestimatedorientationfield,theinputfingerprintimageisadaptiv
- 配套讲稿:
如PPT文件的首页显示word图标,表示该PPT已包含配套word讲稿。双击word图标可打开word文档。
- 特殊限制:
部分文档作品中含有的国旗、国徽等图片,仅作为作品整体效果示例展示,禁止商用。设计者仅对作品中独创性部分享有著作权。
- 关 键 词:
- 文献 翻译 基于 细节 分数 匹配 指纹识别
![提示](https://static.bdocx.com/images/bang_tan.gif)