车牌识别外文文献翻译中英文.docx

车牌识别外文文献翻译中英文.docx

《车牌识别外文文献翻译中英文.docx》由会员分享，可在线阅读，更多相关《车牌识别外文文献翻译中英文.docx（17页珍藏版）》请在冰豆网上搜索。

外文文献翻译（含：

英文原文及中文译文）文献出处：

GaoQ,WangX,XieG.LicensePlateRecognitionBasedOnPriorKnowledgeC/IEEEInternationalConferenceonAutomationandLogistics.IEEE,2007:

2964-2968.英文原文LicensePlateRecognitionBasedOnPriorKnowledgeQianGao,XinnianWangandGongfuXieAbstract-Inthispaper,anewalgorithmbasedonimprovedBP（backpropagation）neuralnetworkforChinesevehiclelicenseplaterecognition（LPR）isdescribed.Theproposedapproachprovidesasolutionforthevehiclelicenseplates（VLP）whichweredegradedseverely.Whatitremarkablydiffersfromthetraditionalmethodsistheapplicationofpriorknowledgeoflicenseplatetotheprocedureoflocation,segmentationandrecognition.Colorcollocationisusedtolocatethelicenseplateintheimage.Dimensionsofeachcharacterareconstant,whichisusedtosegmentthecharacterofVLPs.TheLayoutoftheChineseVLPisanimportantfeature,whichisusedtoconstructaclassifierforrecognizing.Theexperimentalresultsshowthattheimprovedalgorithmiseffectiveundertheconditionthatthelicenseplatesweredegradedseverely.IndexTerms-Licenseplaterecognition,priorknowledge,vehiclelicenseplates,neuralnetwork.I.INTRODUCTIONVehicleLicense-Plate（VLP）recognitionisaveryinterestingbutdifficultproblem.Itisimportantinanumberofapplicationssuchasweight-and-speed-limit,redtrafficinfringement,roadsurveysandparksecurity1.VLPrecognitionsystemconsistsoftheplatelocation,thecharacterssegmentation,andthecharactersrecognition.Thesetasksbecomemoresophisticatedwhendealingwithplateimagestakeninvariousinclinedanglesorundervariouslighting,weatherconditionandcleanlinessoftheplate.Becausethisproblemisusuallyusedinreal-timesystems,itrequiresnotonlyaccuracybutalsofastprocessing.MostexistingVLPrecognitionmethods2,3,4,5reducethecomplexityandincreasetherecognitionratebyusingsomespecificfeaturesoflocalVLPsandestablishingsomeconstrainsontheposition,distancefromthecameratovehicles,andtheinclinedangles.Inaddition,neuralnetworkwasusedtoincreasetherecognitionrate6,7butthetraditionalrecognitionmethodsseldomconsiderthepriorknowledgeofthelocalVLPs.Inthispaper,weproposedanewimprovedlearningmethodofBPalgorithmbasedonspecificfeaturesofChineseVLPs.TheproposedalgorithmovercomesthelowspeedconvergenceofBPneuralnetwork8andremarkableincreasestherecognitionrateespeciallyundertheconditionthatthelicenseplateimagesweredegradeseverely.II.SPECIFICFEATURESOFCHINESEVLPSA.DimensionsAccordingtotheguidelineforvehicleinspection9,alllicenseplatesmustberectangularandhavethedimensionsandhaveall7characterswritteninasingleline.Underpracticalenvironments,thedistancefromthecameratovehiclesandtheinclinedanglesareconstant,soallcharactersofthelicenseplatehaveafixedwidth,andthedistancebetweenthemediumaxesoftwoadjoiningcharactersisfixedandtheratiobetweenwidthandheightisnearlyconstant.Thosefeaturescanbeusedtolocatetheplateandsegmenttheindividualcharacter.B.ColorcollocationoftheplateTherearefourkindsofcolorcollocationfortheChinesevehiclelicenseplate.ThesecolorcollocationsareshownintableI.TABLEIMoreover,militaryvehicleandpolicewagonplatescontainaredcharacterwhichbelongstoaspecificcharacterset.Thisfeaturecanbeusedtoimprovetherecognitionrate.C.LayoutoftheChineseVLPSThecriterionofthevehiclelicenseplatedefinesthecharacterslayoutofChineselicenseplate.AllstandardlicenseplatescontainChinesecharacters,numbersandletterswhichareshowninFig.1.ThefirstoneisaChinesecharacterwhichisanabbreviationofChineseprovinces.ThesecondoneisaletterrangingfromAtoZexcepttheletterI.Thethirdandfourthonesarelettersornumbers.Thefifthtoseventhonesarenumbersrangingfrom0to9only.Howeverthefirstortheseventhonesmayberedcharactersinspecialplates（asshowninFig.1）.Aftersegmentationprocesstheindividualcharacterisextracted.Takingadvantageofthelayoutandcolorcollocationpriorknowledge,theindividualcharacterwillenteroneoftheclasses:

abbreviationsofChineseprovincesset,lettersset,lettersornumbersset,numberset,specialcharactersset.（a）Typicallayout（b）SpecialcharacterFig.1ThelayoutoftheChineselicenseplateIII.THEPROPOSEDALGORITHMThisalgorithmconsistsoffourmodules:

VLPlocation,charactersegmentation,characterclassificationandcharacterrecognition.ThemainstepsoftheflowchartofLPRsystemareshowninFig.2.Firstlythelicenseplateislocatedinaninputimageandcharactersaresegmented.Theneveryindividualcharacterimageenterstheclassifiertodecidewhichclassitbelongsto,andfinallytheBPnetworkdecideswhichcharacterthecharacterimagerepresents.A.Preprocessingthelicenseplate1