畢業(yè)設(shè)計(jì)指紋識(shí)別中若干關(guān)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(shí)代的發(fā)展自動(dòng)指紋識(shí)別技術(shù)已經(jīng)廣泛地應(yīng)用于公安、海關(guān)、銀行、網(wǎng)絡(luò)安全等需要進(jìn)行身份識(shí)別領(lǐng)域。本文概述了自動(dòng)指紋識(shí)別系統(tǒng)的研究現(xiàn)狀和指紋識(shí)別系統(tǒng)的算法流程，并在此基礎(chǔ)上重點(diǎn)研究了指紋圖像的分割算法和指紋圖像細(xì)化算法。首先，本文系統(tǒng)介紹了指紋識(shí)別的研究意義及現(xiàn)狀，并對(duì)指紋識(shí)別原理及實(shí)現(xiàn)過程做了系統(tǒng)介紹。其次，針對(duì)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(shí)別系統(tǒng)的識(shí)別率

2、的問題，本文對(duì)指紋圖像的分割等預(yù)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(duì)指紋圖像的細(xì)化作了深入的研究，使用的細(xì)化模板雖然數(shù)量較多，計(jì)算時(shí)間比較長(zhǎng)，但對(duì)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(xì)節(jié)特征。仿真結(jié)果表明。。。當(dāng)然，鑒于此法速度慢，仍存在毛刺的缺點(diǎn)，還可以有很大的改進(jìn)。最后，研究了指紋識(shí)別過程中特征的提取方法，針對(duì)

3、細(xì)節(jié)點(diǎn)提取過程中存在偽特征點(diǎn)的問題，本文提出采用了一種消除偽特征點(diǎn)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(diǎn)。關(guān)鍵詞：指紋識(shí)別；均值方差；指紋細(xì)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