基于視頻的車流量檢測統(tǒng)計技術(shù)研究.pdf

1、長安大學(xué)碩士學(xué)位論文基于視頻的車流量檢測統(tǒng)計技術(shù)研究姓名：宋俊芳申請學(xué)位級別：碩士專業(yè)：計算機應(yīng)用技術(shù)指導(dǎo)教師：趙祥模20100610AbstractTrafficflowparametersdetectionisallimportantresearchfieldofintelligenttransportationsystemvideobasedtrafficparameterdetectionwiththeadvantagesofr

2、ichinformation，configurationflexibilityandlowcosthasbecomearesearchfocusworldwidelyandhasawideapplicationoutlookThispaperfocusesonvehicledetectionandstatisticsusingvideoprocessingmethodThemaincontentsinclude：(1)Detection

3、regionsettingandgeometriccorrectionInthefirstflameofthevideodatathelaneissetmanuallyandthelaneinformationisstored；、析tllthelaneboundaryinformationdetectionareaissetverticallytotrafficdirection；theimageindetectionareaispro

4、jectedtoarectangularareathroughgeometrictransformation(2)BinarizationsegmentationofmovingobjectTraditionalmethodsoftargetdetectionarecomparedandanalyzedAgainstfeaturesofactualtrafficscenes，themovingtargetdetectionalgorit

5、hmiscombiningthemethodsofframedifferenceandbackgrounddifference，isimplementedThispartmainlyimprovethealgorithmofframedifference，backgroundupdating，andemptyandvirtualshadoweliminating(3)TargetextractionandcountingVehiclei

6、nformationdataareextractedfromthebinaryimageindetectionregionthetwo—dimensionalboundaryoftargetregionisdetermined，andaccordingtocenterofgravitylocationofthetargetareaboundaryappropriatematchingcriterionischosentodetermin

7、evehiclesandcountthenumberofvehiclesThispaperimplementsareliableandrealtimevehiclesdetectionandstatistics，whichCallbeusedtealizetheprecisetargetdetectionandvehiclesstatisticsShanghaiouterringmeasureddatashowthat：thealgor

8、ithmrelatedtothispaperCanaccuratelymeasurethetrafficflowovertime，andtheactualmeasurementaccuracyattains95％Experimentalresultsshowthatthemethodoftrafficflowdetectionandstatisticsmeettherequirementsofvehicleflowdetectionan