畢業(yè)設計指紋識別中若干關鍵問題研究

1、Comment[zyf1]:摘要還可以再修改一下。摘要隨著信息時代的發(fā)展自動指紋識別技術已經(jīng)廣泛地應用于公安、海關、銀行、網(wǎng)絡安全等需要進行身份識別領域。本文概述了自動指紋識別系統(tǒng)的研究現(xiàn)狀和指紋識別系統(tǒng)的算法流程，并在此基礎上重點研究了指紋圖像的分割算法和指紋圖像細化算法。首先，本文系統(tǒng)介紹了指紋識別的研究意義及現(xiàn)狀，并對指紋識別原理及實現(xiàn)過程做了系統(tǒng)介紹。其次，針對在指紋采集過程中存在著大量的低質(zhì)量指紋圖像，影響指紋識別系統(tǒng)的識別率

2、的問題，本文對指紋圖像的分割等預處理作了較為深入的研究，提出采用了均值方差的指紋圖像分割算法，仿真結(jié)果表明，該方法有效的地改善了指紋圖像的質(zhì)量。此外，對指紋圖像的細化作了深入的研究，使用的細化模板雖然數(shù)量較多，計算時間比較長，但對于像素的處理很到位，使得毛刺比較少，保持原圖像的基本結(jié)構(gòu)特性，還最大地保留了細節(jié)特征。仿真結(jié)果表明。。。當然，鑒于此法速度慢，仍存在毛刺的缺點，還可以有很大的改進。最后，研究了指紋識別過程中特征的提取方法，針對

3、細節(jié)點提取過程中存在偽特征點的問題，本文提出采用了一種消除偽特征點的算法。仿真結(jié)果表明，這種算法可以有效的去除偽特征點。關鍵詞：指紋識別；均值方差；指紋細化；特征提取ABSTRACTWiththedevelopmentsoftheinfmationageautomatedfingerprintidentificationtechnologyhasbeenwidelyusedinpublicsecuritycustomsbanking.T

4、hisarticleoutlinesboththestudystatusalgithmprocessofaAutomatedFingerprintRecognitionSystemonthisbasisfocusesonthesegmentationoffingerprintimagefingerprintimagethinningalgithmtodothestudy.Firstthepaperdescribestheresearch

5、significanceoffingerprintrecognitionitsstatus.Fingerprintrecognitiontheythekeystepsinthealgithmareintroducedinthispart.Theyarealsousedfprogrammingverification.Secondlyinthefingerprintcollectionprocessthereareexistencesof

6、alargenumberoflowqualityfingerprintimagesaffectingtheidentificationrateoffingerprintrecognitionsystemproblems.Thepaperfocusesonfingerprintimagesegmentationwhichbelongstopreresearchmakesamethoughstudyputtingfwardafingerpr

7、intimagesegmentationalgithmbyusedthemeanvarianceoftheimage.Calculatingthemeanvarianceofeachpiecethenifthisresultisalmostcloseto0thatitisregardedasthebackground.Theareavarianceisnotzerowillusethresholdsegmentationalgithm.

8、Inadditionthefingerprintimagethinningisstudieddeeplyalthoughthenumberoftemplatestouseisalotcalculationtimeislongerthedealisinplacefthepixelsmakingrelativelyfewglitchesmaintainthebasicstructuralacteristicsoftheiginalimage

9、retainingthedetailsofthebiggestfeaturestoo.Ofcoursegiventheslowspeedofthismethodtherearestillshtcomingsofburrcanbeusedfmakingagreatimprovement.Finallythefeatureextractionmethodinthefingerprintrecognitionprocessisstudied.

10、Focusonfeatureextractionmethodsfextractionofminutiaefeaturepointsexistinpseudoproblemweproposeapseudofeaturepointstoeliminatethealgithm.Simulationresultsshowthatthealgithmcaneffectivelyremovepseudofeaturepoints.Keywds:Ke