基于圖像處理與神經(jīng)網(wǎng)絡(luò)的車牌識別技術(shù)的研究.pdf

1、，r‘跫裹I●，’■爭隧曩寮代號分類號10701邢914學號罌!螂!衛(wèi)!密級公開面簧它矛料輟兜警碩士學位論文題(中、英文)目萎王阻像處疆與麴經(jīng)圈蟹監(jiān)皇牌遲到繭鲞自魚研寶一量；艘喲h旦n__c隨鯉墅險；蠼孕韭；她照址!觸擔P臻蘭曼一險0要銎韭；盟盤i地Ⅲ曼墼!d旦n_～IEage_P__reProcess堡g衛(wèi)照舅嬰!亟一№；嬰地一作者姓名墨囂燕指導教師姓名、職務(wù)j肇良盔一劇熬援學科門裝互望學科、專業(yè)，變遙焦宣王強叢控劍。提交論文霞飄一=

2、暑旦Q圭壘=盈一一AbstractAbstractWiththefantasticspurineconomyandrapiddevelopmentoftheowningamountofautomobile，thehighwaycommunicationbecomesoneofthemostimportantcommunicationsandtransportationwaysinourcountry，butthetraditionalt

3、ransportationpracticaldemandsnowadays，alsocrowdedtrafficneedsmofeadvancedandmoreeffectivetrafficadministrationandsystemSoIntelligentTransportationSystems(ITS)whichmakesuseofimagepreprocessingtechnologyandcommunicationtec

4、hnologybecomesmaindirectionoftrafficadministrationLicensePlateRecognition(LPR)isoneofthecriticaltechniquesandthekeyofITS。Thepaperisbasedonimagepreprocessingtechnology、licenseplatelocationtechnology、charactersegmentationt

5、echnologyandBPneuralnetworktechnologytostudyLPRFirst，throughgrayequalization、medianfilterandsoon，thenoisewillbereducedwellandthelengthways—textureareacanbestrengthened；throughdeeplystudyinlicenseplatecharacterandlicensep

6、latelocation，alicenseplatelocationalgorithmbasedonhorizontalSCanandverticalpr鉚ectionisputforwardthatcansearchthelicenseplateareaaccurately；basedonstructurecharacteristicandtranscendentinformationoflicenseplate，characters

7、egmentationwillbedonewell；last，thepaperanalyzesthestructureoftheneuralnetworkwithrelevanttheoriesandemphaticallydiscussesthetheoryprincipleoftheneuralnetworkandtheapplicationmethodincharacterrecognitionwithitAsthetestres

8、ult，itprovesthatthisrecognitionsystemcanrelativelyaccuratetolocatelicenseplateandrecognizecharactersandsystemsperformanceisgoodItshowsthatcombinedpretreatmentsandrecognitiontechniquescanimprovetheabilityofrecognition，and

9、thecombinationofneuralnetworkandartificialintelligenceundertheprincipleofefficiencyandpracticalitywillbetwomajordevelopmenttendenciesofpatternrecognitionKeyword：Licenseplaterecognition，Imagesegmentation，Opticalcharacterr