文档下载 文档下载导航
设为首页 | 加入收藏
搜索 请输入内容:  
 导航当前位置: 文档下载 > 所有分类 > Processing Shape, Motion and Three-dimensional Shape-from-motion in
侵权投诉

Processing Shape, Motion and Three-dimensional Shape-from-motion in

Shape and motion are complementary visual features and each appears to be processed in unique cortical areas. However, object motion is a powerful cue for the perception of three-dimensional (3-D) shape, implying that the two types of information — motion

whichthreetypesofstimuliwerepresented:(i)randomlymovingdots,(ii)flickeringdots,and(iii)stationarydots.Eachblocklasted20sandthereweresixpresentationsofeachcondition.Therandomlymovingdotsconsistedofanarrayof450dotssubtending10°ofvisualangle.Eachdot moved in arandom direction withconstant speed (4°/s).Themotiondisplaysweregeneratedbycreatingmoviesequenceswitheachframestoredasaseparatebitmap.Thesequenceplayedforwardthenreversedsothatrandommotionwascontinuallyobserved.Topromoteeyefixationandtoequateattentionaldemandsacrossconditions,subjectswereaskedtodetectrandomlyoccurring,brief(60ms)luminanceincreasesofthefixationcross.Forthebaselinestationarycondition,arandomlychosenframefromthemoviesequencewaspresentedalongwiththesamefixation luminance-detection task.

Toseparateresponsesduepurelytotemporalaspectsofthemotionstimulus,signalassociatedwithaflickeringdisplaywascomparedtothatproducedbyrandommotionstimuli.Theuseofflickermaintainstemporalpropertiesbutdoesnothavethesamespatio-temporalstructureofmotion.Flickerstimuliweregeneratedbyrandomizingtheorderofthemovieframesusedforthemotionstimuliandwerepresentedwiththesamefixation luminance-detection task.

Shape

Toidentifyshape-processingregions,threeconditionswerepresentedovertwoscans:(i)randomlines,(ii)linesthatformed2-Dshapesand(iii)lines thatformed 3-Dshapes(Fig. 1).(Thefirst two subjectsscanneddidnotviewthe2-Dshapecondition.)Eachconditionwaspresentedin30sblocksand presenteda total ofeight timesoverthetwoscans. Tocontrolforpotentialdifferencesinattentionsubjectsrespondedwithabuttonpress tobrief(100 ms),randomlyoccurringgloballuminance reductions(40%)inthestimuliinallconditions.The2-Dshapesweregeneratedbyrandomlyselecting4–7 verticesataminimumdistancefromfixationandconnectingthevertices.Therandomlineswerecreatedbybreakingthe2-Dshapesattheirintersectionsandrandomlyrepositioningthemwhilemaintainingthesamemeandistance(4°)fromfixation.The3-Dshapeswere the sameas the2-Dshapes in allrespects withtheadditionofsmallextensionsthataddedperceiveddepth(Fig.1).Eachfigurewaspresentedfor 750 ms.

Shape-from-motion

Random-dotdisplayswerepresentedunderthreeconditions:(i)stationary dots,(ii) ‘velocity-scrambled’ movingdotsand(iii) projectionsofarandom-dotpatternontomovingthree-dimensionalgeometricshapes(shape-from-motion,SFM)(Fig.2).Conditionswerepresentedin30sblocksandpresentedatotalofeighttimesovertwoimagingscans.Thestimuliineachconditioncontained450dotsandsubtended10°.Thedotsintheshape-from-motion(SFM)stimuliwereprojectionsofrigid,trans-parent,geometricshapesincludingacube,cylinderand‘house-shaped’figures.Thesethreeshapeswereusedbecausetheyhadlimitedconfoundingshapecues.ManySFMshapes(e.g.pyramid)haveadditionalshapecuescreatedbytheborderofthestimuli(triangle).Dotswererandomlyselectedfromauniformdistributiononthe objectsurfaceandkeptfixedrelativetotherotatingobjectsurfaceandorthographicallyprojectedontotheimageplane.Thedotswererotatedonarandomlychosen 3-Daxisfor40° in1.5° increments.Thedotsdidnothavelimitedlifetimes. Adot onthefront surface ofan object moving in onedirectioneventuallyappearedonthebacksurfacemovingintheoppositedirection.Eachstimuluspresentationlasted890msfollowedby110msblank-screendelaybeforethenextstimuluspresentation.TheperceptionofshapeintheSFMstimuliiscreatedbythedistribution ofrelativespeedthatisperceivedasdifferingdepthsby the visualsystem(motionparallax).Examplesofthe SFMstimuliandthecontrolstimuli (describedbelow) can beviewedatdoc.guandang.net.

TheSFM stimuli were compared tovelocity-scrambledstimuli–thesehadthesamevelocities(directionandspeed)astheSFMstimulibutwererandomlyreassignedtodifferentdotscreatingperceptuallyrandommotion,similartoSiegelandAndersen(SiegelandAndersen,1988).ThefirstframefromaSFMstimulusalsoservedasthefirstframeinthecorrespondingvelocity-scrambled condition. Thebaseline,stationarydotconditionpresentedarandomlychosenframefromtheSFMstimuli.As

Figureprocessing.2.ExamplesActualfiguresofthewerestimuliwhiteuseddotstoonanalyzeablackstructure-from-motionbackgroundsubtending(SFM)9°ofvisualangle.Thevelocity-scrambled stimulihadidenticalvelocitiesas the SFMstimulibutwererandomlyre-assignedtodifferentdotscreatingperceptuallyrandommotion.(Thedashedlineandarrowsareaddedforillustration.)Moviedemonstrationsofthestimulicanbeviewedatdoc.guandang.net.

withthelinedrawingstimuli,subjectswereaskedto detectbrief(100ms),randomlyoccurringluminancereductions(40%)inthestimuli.MRIAcquisition

Foralltheexperiments,scanningwasdoneona1.5TMarconiscannerattheNorthernCaliforniaVeteransAdministrationHealthCareSysteminMartinez,California.Standardimagingprocedures(TE=40ms;flipangle =90°,TR= 2000ms) wereused. Datawerecollectedfrom205mmthickaxialslices(128×128,FOV=240mm)that,formostsubjects,coveredtheentirebrain.Toaidinvisualizationofstatisticalmapsandtomoreeasilydefinevisualareas,datawereprojectedontoinflatedcorticesusingFreeSurfer(DaleandSereno,1993;Daleetal.,1999;Fischletal.,1999,2001)usinghigh-resolutionspin-echoanatomicalscanscollectedfromeachsubject(FOV=240mm,matrix=256×212×256,TE= 4.47ms,TR= 15 ms,flipangle = 35).

Data Analysis

Thedatawerepre-processedusingSPM99(doc.guandang.net),includingrealignmentoffunctionalscanstocorrectforheadmotion,normalizationintostandardizedcoordinates,andspatialsmoothing(4 mm3FWHMkernel).StatisticalmapsweregeneratedforeachsubjectusingSPM99andservedasabasisfordefiningregions-of-interest (ROIs).Inalloftheanalysesdescribedbelow,unlessotherwisementioned,anROIwasdefinedasacontiguousregionofvoxelssurpassingaminimumstatisticalthresholdofP< 10–3.Tocalculatepercentchange,wefirstderivedthelinearlydetrendedtimecourseoftheMRsignalintensityforeachvoxelinanROI.Theaveragepercentsignalchangewasthencalculatedforeachstimulustype,relativetotheaveragesignalintensityduringabaselinecondition.Baselineconditionsdifferedfordifferentanalysesandaredescribedbelow.BecausethefMRIresponsetypicallylags4–6saftertheneuralresponse,thefirstimageineachepochwasassignedtotheconditionoftheprecedingepochandweomittedthenexttwoimagesfromanalysis(i.e.ignoredtransitionimagesbetweenepochs). Sincenohemisphericdifferences wereobserved, ifanROIwaspresentinbothhemispheres(e.g.leftandrightMT+),thedatawereaveragedtogether.Pairedtwo-tailedt-testswereusedtocomparepercentchangedatabetweenconditions.NocorrectionformultiplevoxelwisecomparisonswasusedbecausethedatawereanalyzedwithinROIsthathad beendefined.

MotionAnalysis

MotionROIsweredefinedbyacomparisonofactivationsinstationaryandrandommovingdotconditions.Becausemanyvisualareas

Processing Shape, Motion and Three-dimensional Shape-from-motion in

Processing Shape, Motion and Three-dimensional Shape-from-motion in

are

CerebralCortex May2003,V 13N5509

上一页第2页

你可能喜欢

热门文档

相关文档

站点地图 | 文档上传 | 侵权投诉 | 手机版
本站所有资源均来自互联网,本站只负责收集和整理,均不承担任何法律责任,如有侵权等其它行为请联系我们.
文档下载 Copyright 2013 doc.guandang.net All Rights Reserved.  -  email