基于视频的车型自动识别系统的实现

《基于视频的车型自动识别系统的实现》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。

1、郑州大学硕士学位论文基于视频的车型自动识别系统的实现姓名：范伊红申请学位级别：硕士专业：测试计量技术及仪器指导教师：吕运朋;张元20060520郑州大学硕士学位论文AbstractThemesisissupportedbytheIIlllovationFoundationofExcellentIntellectuaIinHenanProvince．TbdatemtelIigent1hnsponationsystem(ITs)hasbecomeoneofthemostpopularworldwideresearchfieldsintramctechnologyTheres

2、earchtopicsacknowledgedintemationallyincludeadvallcedtmmcmanagementsystem'advancedtmmcinfb珊ationservicesystem，commercialvehicletransportationandmanagementsystem，electmnictoUsystem，public仃aIlsportationmanagementsystem，eme唱encymanagementsystem，andadvancedcarmonitoringsystem．Asar印resentatio

3、nofelec仃onicautomatictollsystem，theNon—StoppingElec仃onicTbllSystem(NSETCS)baseduponnon—contactingIntegratecircuit(Ic)cardisausefIllmetllodtosolvethisproblembecauseitcaneⅡ色ctivelyimpmvethetra皿sponationandscrviceofhighways．SincecarmodelisⅡleessentialbasetodeteminethehigbwayt01l，CarModelIde

4、nt讯cationsystem(CMIS)isanimportantpartofTheNSETCSbasedonnon-contactingICcard，whichnormallyadoptsinfonllation如siontecllllologytomakeadecision．ThesystemnsestheinfbⅡnationoftheCarModelexponedbyCMIS，licenseplateexportedbyLicensePlateReco舭itionSystem，atldthedatasavedinthenon—contactingICcardt

5、oobtaintolldecreeandhencegreatlyimpmvesthereliabilityofthesystem．Thetllesisprcsentedareal—tilneCMISbasedonthevideoimageacquired矗omvideocamerasbyemployingmoVingobjectdetectionalgorithIn，featureex廿actionandpattemrecognitionalgorithm．ThemainworksinmismesisaresunHnarizedasfbllows：1．Desi印edan

6、dimplementedavideoImagePmcessingSystembasedonDirectShow，a11dpresentedthekemelso脚arecode．2．stIldiedtheeXistingdigitalimageprocessingtecllllologyaIldalgorjthms，andappliedthemtoCarModelReco弘itionSystemtosuppressallkindsofnoise．3．Reviewed也emainmovingobjectdetectionalgorithmsandtheiradvantage

7、sanddisadVantages．DeVelopedanewalgorithmforthedetectionofhi曲-speedmovingobjectsbasedonpictIlreblocksandHVScolorspaceofimagedia'crencestechniques．4．Featllreex缸action．Sevencarim刮geinV撕antmom肋tsareextracted仔omthecarHI郑州大学碗十学位论文AbstractoutIineexc印tfort11enomalfeaturessuchasth