基于二維灰度圖像的分區(qū)域三維表面重建技術(shù)研究.pdf

1、武漢理工大學(xué)碩士學(xué)位論文基于二維灰度圖像的分區(qū)域三維表面重建技術(shù)研究姓名：張忭申請(qǐng)學(xué)位級(jí)別：碩士專業(yè)：計(jì)算機(jī)應(yīng)用技術(shù)指導(dǎo)教師：徐東平20080501武漢理工大學(xué)碩十學(xué)位論文AbstractThetechniquewhich3Dsurfaceshapeisrecoveredfrom2Dgrayimage(shapefromshading，SFS)isoneofthekeytechniquesof3Dshaperecoveryincompu

2、terfieldDuetoitsfeature，whichcanrestore3Dshapeinformationjustfromonegray—levelimage，recentlySFSgetsgreatimprovedanddevelopsrapidlyinmethodologyandapplicationandbecomeanewresearchof3DpassivereconstructionAtthebeginningoft

3、histhesis，acomprehensivesurveyaboutSFSresearchworkisgiven，andtherecentresearchfocusesinthisfieldaresummarizedForconsiderationofprincipleofphotometrytheilluminationofthreelocalilluminationmodelsthatareLambert，PhongandOren

4、—Nayarmodel，aredeterminedAndwithSFSanalyticalprinciple，thecorrectilluminationmodelforSFSisfixedalreadyOnthebaseofunderstandingforSFSprobleminnature，asystematicsummaryismadeforallcurrentSFSalgorithmsatpresent，whicharemini

5、mizationapproaches，propagationapproaches，localapproaches，andlinearapproachesMeanwhile，basedonprincipleofphotometryandtheoryofnumericalsolutionfornonlinearpartialdifferentialequation，thecorrespondingalgorithmbasedontheLam

6、bertreflectancemodelforminimizationapproachisprovidedMeanwhile，analgorithmofreversecalculationinthespotgrayofvirtualspacebygivenheightandgradienthasalsobeenpresented，whichisinordertoprovideeachpixelarealcolorandgetanimag

7、ewiththestrongestfeelingofrealityrealityBythiswaythecomplexityanderrorbecauseofthecameracalibrationcallbeavoidedandaperfectresearchflatwillbeacquiredAtleast，withtheresultofexperiment，whichshowsthatthealgorithmisfeasiblea