摘要Subject: TheResearchofTrafficFlowDetectionAlgorithm BasedonMachineVision Specialty: VehicleEngineering Name: WuGuo-wei Supervisor: GuoZhi-qiang XieJin-fa ABSTRACTThestudyofTrafficFlowon-linedetectionisbasedonmachinevision,,thedetectionoftrafficflowmethodexistsintheprocessofdetectionoflowefficiency,highrateoffalsedetections,slowspeedandsoondoesn’tmeetrequirementsofreal-,thealgorithmstudyoftrafficflowdetectionbasedonmachinevisionisfocusedonautomotiveimageprocessingandautoimagefeatureextraction,autoimagetargetdetection,,firstofall,thevehicle'simagesthatarecollectedwillbechangedfor24-bitGDIbitmapbyA/D,thevehicle’simagesconvertedareprocessedbysmoothing,segmentation,gray,,afterthevehicleimageextractionprocessingsizeofthearea,perimeterandcentroidoftheregionandotherparameters,basedontheextractedcharacteristicparameterstoachieveimagetargetdetection,andstatisticalvehiculartraffic,,usingdifferential-basedmethodtodetectcarimagetarget,andcalculatethetrafficflow,uracyrateof98%.icalgorithmasajudgingdevice,work,work,workwhichoptimizedbyGA,pletingtraffickflow'sdetecting,thecorrectingrateofjudgementismorethan96%.Thefinaldesignistrafficflowinspectionsystemwhichbasedonmachinevision,,theTrafficFlowDetectionsystemisdesignedinthispaper, andtheII摘要experimentsverifythatthissystemmeetsrequirementsofreal-timeandthedevelopmentofIntelligentTransportation,atthesametime,feasibilityandcorrectnessofthispaper’:Machinevision,Imageprocessing,Flowdetection,works,FeatureparametersDissertationType:,交通需求量越来越大、交通拥堵情况日益严重、交通事故频繁发生,交通问题正日益成为当今世界各国面临的共同问题。就我国而言,人口数量的不断增长,经济
基于机器视觉的汽车流量检测算法研究-车辆工程专业论文 来自淘豆网m.daumloan.com转载请标明出处.
