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

1、，r‘跫裹I●，’■爭(zhēng)隧曩寮代號(hào)分類號(hào)10701邢914學(xué)號(hào)罌!螂!衛(wèi)!密級(jí)公開面簧它矛料輟兜警碩士學(xué)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(xiǎn)；蠼孕韭；她照址!觸擔(dān)P臻蘭曼一險(xiǎn)0要銎韭；盟盤i地Ⅲ曼墼!d旦n_～I(xiàn)Eage_P__reProcess堡g衛(wèi)照舅嬰!亟一№；嬰地一作者姓名墨囂燕指導(dǎo)教師姓名、職務(wù)j肇良盔一劇熬援學(xué)科門裝互望學(xué)科、專業(yè)，變遙焦宣王強(qiáng)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