基于双目立体全景视觉传感器的道路宽度实时检测技术的分析

《基于双目立体全景视觉传感器的道路宽度实时检测技术的分析》由会员上传分享，免费在线阅读，更多相关内容在行业资料-天天文库。

1、浙江工业大学硕士学位论文理引入双目立体全景视觉领域，实现三维球体空间内任意物点的空间坐标测量；进而，根据道路应用环境，建立空间成像模型实现道路宽度的计算。最后，对测量的准确性和实时性进行总结，分析各影响因素对测量准确度的影响，并提出改进方法。关键词：全方位视觉传感器(0DVS)，移动道路测量系统，双目立体全景视觉系统，双目测距，SIFT／ASIFT浙江工业大学硕士学位论文TheResearchofReal．timeRoadWidthMeasurementTechnologyBasedonBinocularStere0rano

2、ramlCVlSlonSenS0r，’一J1n●I7●●』1ABSTRACTWimthedeVelopmentofChineseurb锄ization，thepressureofpopulation，ene唱y，胁sportation，inf．mstnJcture锄dothersisgraduallyincreased．InordertopromotethedeVelopmentofurbancons仃uction，tlleconceptof”DigitalCit)r”wasproposed．”DigitalCi够”isⅡl

3、e恤ee·dimensionaldescriptiononci哆bymulti—resolution，multi—scale，multi—dimensionalspace—timeandmulti·species，stylet0achieVedigitaleconomydeVelopmentandpublicserVice．Thecons仃1Jctionof劬d锄entaldatabaseisthes仃ategicfoundationofI-digitalcity”，andmedatacollection锄dprocessi

4、ngistlleprerequisiteonit．HoweVer't11e仃aditionalwofIdngpa钍em，notonlyrequiresalotofm锄power'material锄dfin锄cials，butalsoine衢cient．Itisdi伍cultt0keepupwimr印idCh锄gesinurb锄n锄sportationcons劬ction．TheMobileM印pingSystem(MobileMappingSystem，MMS)proVidesaconVenient锄deconomicpat

5、tem，WhichcanmeettIlerequirementsofinf}aus仃uctureda眦Iasecons廿．uctioninintelligencen．anspor眦ion．TheMMSiscutting-edgetecllnologyinm印pingcu玎l；ntly．nis矗ttedint11eVehicleglobalpositioningsystem(GPS)，Videosystem(CCD)，inertialnaVigationsyStem(INS)ortlledeadreckoningsystem(

6、DR)a11do廿leradVancedsensors卸dequipment，嬲theVehiclesmoVingwimhigh—speed，t11eM[MScanacquirethespatialdataandattributedataofroads，allthese讹arestoredintlleonboardcomputerinrealtimeorViaprocessingtof0册aVariet)rofusefhldatasetsinte伊纰edsystem．HowtousetheCCDVideosystemtoca

7、pturerichimage行ombothsidesoft11eroads卸dsurfacef．eanⅡ℃sortocalculatespatialda‰砌butedataaccuratelyin浙江工业大学硕士学位论文reaI-timetoachieVerealthree—dimensionalinfomationsystem，thisisakeyissueoft11eMMS．nisselectedarticleisbasedontheVideosystemofbinocularstereopanoramicvisions

8、ensoronwidthmeaLsurement印plicationoftheroad．ThebinocularstereopanoramicVisionsensorisdesi印edVerticallyby铆oOmidirectionalVisionsensors，“sdesigllc锄