简介:Stereomatchingisafundamentalandcrucialproblemincomputervision.Inthelastdecades,manyresearchershavebeenworkingonitandmadegreatprogress.Generallystereoalgorithmscanbeclassifiedintolocalmethodsandglobalmethods.Inthispaper,thechallengesofstereomatchingarefirstintroduced,andthenwefocusonlocalapproacheswhichhavesimplerstructuresandhigherefficiencythanglobalones.Localalgorithmsgenerallyperformfoursteps:costcomputation,costaggregation,disparitycomputationanddisparityrefinement.Everystepisdeeplyinvestigated,andmostworkfocusesoncostaggregation.Westudiedmostoftheclassicallocalmethodsanddividethemintoseveralclasses.Theclassificationwellillustratesthedevelopmenthistoryoflocalstereocorrespondenceandshowstheessenceoflocalmatchingalongwithitsimportantanddifficultpoints.Attheendwegivethefuturedevelopmenttrendoflocalmethods.
简介:AMULTIPLESTRUCTURERECONSTRUCTIONALGORITHMBASEDONWIREFRAMEMODELDuJiangchuanPaulW.F.PoonAbstractA“LAG&RLC”visualizationalgorith...
简介:Opticalmotioncaptureisanincreasinglypopularanimationtechnique.Inthelastfewyears,plentyofmethodshavebeenproposedforkey-frameextractionofmotioncapturedata,anditisacommonmethodtoextractkey-frameusingquaternion.Here,onemaindifficultyisduetothefactthatpreviousalgorithmsoftenneedtomanuallysetvariousparameters.Inaddition,itisproblematictopredefinetheappropriatethresholdwithoutknowingthedatacontent.Inthispaper,wepresentanoveladaptivethreshold-basedextractionmethod.Key-framecanbefoundaccordingtoquaterniondistance.Weproposeasimpleandefficientalgorithmtoextractkey-framefromamotionsequencebasedonadaptivethreshold.Itisconvenientwithnoneedtopredefineparameterstomeetcertaincompressionratio.Experimentalresultsofmanymotioncaptureswithdifferenttraitsdemonstrategoodperformanceoftheproposedalgorithm.Ourexperimentsshowthatonecantypicallycutdowntheprocessofextractionfromseveralminutestoacoupleofseconds.
简介:Thereportmainlystudiedthecrawlerframemotionplatformtoreduceweightandincreasetheintensity.ReportdescribedthestructuraldesignprocesswhichusingCAD/CAEtechnologyforsolidmodeling,simulationanalysis,parameteroptimization.AnditalsoexplainedtheoutstandingadvantagesofCAD/CAEtechnologyinmechanicaldesignaswellassimulationanalysis.