基于多線激光條紋測量的匹配與重建研究

1、青島大學碩士學位論文基于多線激光條紋測量的匹配與重建研究姓名：劉文學申請學位級別：碩士專業(yè)：計算機應(yīng)用技術(shù)指導教師：方漪20070601AbstractThepaperbroughtforwardavisioncoordinatemeasuringtechnologybasedonthemulti—stripelasermarksInthistechnologycorrespondencebetweenmarkswaseffective

2、lycompletedbyepipolarconstraintandthen3DobjectwasreconstructedThistechnologyimprovedmeasuringmethodattachingmarksonobject，overcameshortcomingsofbeingunstableandlowefficiencythatmanualmeasuringmeanshad，anditwasanon—contac

3、tmeasuringmethodcapturingfigureof3DobjectincelerityFirstlyinthemethod，imageswithaberrantlaserstripescontaininginformationaboutfigureof3Dobjectwerecaptured，thenimprovedsubtractionoftheimageswasusedtodealwiththeseimages，th

4、enpretreatment，thresholddivisionandstripethinningwereperformedonthemSecondlyfurtheranalysiswasperformedbasedonmathematicsmodelofbinocular3Dvisionandimagesprocessingarithmetic，inwhichthematchingbetweenstripeswasfulfilledo

5、nstrategyofepipolarconstraintcorrespondenceThenthedataofobject3DfigurewasobtainedThirdlyexperimentswerecarriedouttovalidatetheaccuracyandvalidityofthemethodofepipolarconstraintByOpenGL，resultsofexperimentsweredisplayedan

6、dpartofwhichwereexecutedbydotcloudprocessingResearchresultscanbedescribedconcretelyasfollowing：1、TheproductionofstripedlasermarksThelasermarksgeneratorshouldbedesignedwithproperfocallength，intensityandcombiningmeansAfter

7、myriadsofexperiments，aproperlasermarkgeneratorwasdesigned，andseveralimportantparameterswereexactlydescribedintheend2、CapturingandprocessingofmarkedimagesInordertoextractmultistripelasermarksfromimages，thispaperpresentedp

8、roperimagecapturingprojectTheoperationofimprovedsubtractionofimageswasusedtoextractthestripelasermarks，whicheffectivelysolvedtheeliminationofnoiseinterferenceproducedbyunstablecapturingconditionsfromlightandcamerasForthe

9、simplicityofstructureoflaserstripesimages，theOstumethodwaseffectivelyappliedinimagedivisionThensinglepixelstripesweregeneratedviathinningmeansofmorphologicprocessingofimage3、Correspondencebasedonepipolarconstraintandrebu