車輛視覺導(dǎo)航方法研究.pdf

1、西北工業(yè)大學(xué)碩士學(xué)位論文車輛視覺導(dǎo)航方法研究姓名：盧衛(wèi)娜申請學(xué)位級別：碩士專業(yè)：系統(tǒng)工程指導(dǎo)教師：史忠科20060301pq北1業(yè)人學(xué)碩十。產(chǎn)位論文AbstractAbstractIntelligentvehicletechniqueisapopresearchareaofITS(IntelligentTransportationSystem)，andtheresearchonroadrecognitionwhichisakeytech

2、niqneIbrintelligentvehiclenavigationsystem，doesgreatcontributiontosafedrivingandevenunmanneddrivingInthisthesis，topreventanintelligentvehiclefromdepartingthelaneinthevisionbasednavigation，anintegratedmethodbasedonmonocul

3、arvisionisproposedtodetectthelanemarkingandroadboundaryinfrontofthecarTherelativepositionanddirectionofthelaneandvehiclearethenacquiredDuringthegeneraldesigningofavision—basednavigationsystem，thesoftwareprocessingalgorit

4、hmispresentedasbelowFirstly，thecolorgraytransform，imagesmoothingandcontrastenhancementalgorithmsareappliedtotheimagepreproeessingThroughtheprojectionmethod，theleveldemarcationlinebetweentheroadsurfaceandthebackgrotmdisfo

5、undout，andthentheregionofinterestisdefined，Secondlythecharactersofthelanemarkingandroadboundaryareextractedwiththehelpofthemethodslikegradientoperatorconvolution，gradientanglequantization，non—maximumsuppressionandedgepoi

6、ntsearchingFortheparticularbrightnessofthelanemarking，all0’ISUmethodisalsoappliedIothelanemarkingcharactersextractionThirdlythestraightandcrookedroadmodelsarebuiltaccordingtopeople’sdrivingexperience，priorknowledgeandmat

7、hematicdescriptionThenthecharacteristicpointsofthelanemarkingandroadboundaryarematchedtothecorrespondingmodelbyleast—squaresfitInthiswaythepositionandshapeoftheroadareidentifiedWhat’smore，thereferencepathfordriving，theac

8、tualpositiondeviationandangledeviationofthedrivingcararestudiedinthethesisAcircularprocessingmethodisdesignedformassimagesequencesusingtheresultsofformerflameteducethesearchingrange，sothecalculatingefficiencyisincreasedt

9、oagreatextentMostofthealgorithmsinthepaperhavebeenvalidatedbythevideoscapturedfromrealtralHcroadscenes，andtheexperimentalresultsshowthatthemethodsareefllcientstableandaccurateKeywords：Intelligentvehicle，Visionbasednaviga