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 > IT/计算机 > paper1220_final
侵权投诉

paper1220_final

快速Mean-shift聚类

takesonlytotal0.958secondsonCPUtoperformmeanshiftclusteringwithsamplingfactorn/m=1,024.

InFigure4,wepresentimagesegmentationresultsfortheimageswithdifferentsizes,andgivethecomparisonre-sultswithstandardmeanshiftmethod[CM02],theaccel-eratedmethodofParisandDurand[PD07],andcompactlyrepresentedKDEofFreedmanandKisilev[FK09].Wecom-prisewiththesemethodsonbothperformanceandsegmen-tationquality.Comparedwith[CM02],someweakfeaturesmaybelostusingourmethodsincetheymaybeincorpo-ratedintothesalientfeaturesduringdataGaussianKD-treeclustering,however,thesalientfeaturesmaybebetterkept,asillustratedinFigure4.AsshowninFigure4,giventhesamesamplingfactor,ourmethodgenerateshigherqualitycomparedwith[PD07]and[FK09],especiallyattheregionswithweakedges.Thecomplexityof[PD07]dependsonthedimensiondofthepoint,whenprocessinghigh-dimensionaldata,thismethoddoesnotshowmuchadvantage.However,ourmethodisfastevenwithlowsamplingfactorandhighdi-mensionaldatasets.asshowninTable1.Ittakesourmethod5.91secondtocluster6.6×106pixelsonCPU,whileittake105.5secondsusing[PD07].Usingourmethodincorpo-ratingGPUimplementation,ourmethodshowsgreaterad-vantagewhenprocessinglargedatasetswithhighdimen-sionalfeaturevector,ittakeslessthan0.2secondtocluster6.6×106pixelsonGPU.

4.2.Videosegmentation

Meanshiftclusteringcanalsobeusedinvideosegmen-tation[DeM02].Asvideostreamingusuallycontainsmil-lionsofpixels,practicalvideosegmentationusingmean

shiftclusteringdependsheavilyontheperformanceofthemeanshiftprocedure.Inaddition,comparedwithimagedata,thedimensionsoffeaturespacearehigher,whichfur-therincreasethecomputationalcomplexityofmeanshiftprocedure.Thus,itisimpracticabletosegmentlongrangevideostreamingbyperformingstandardmeanshiftcluster-ingwithoutusingaccelerationtechniques.However,usingourfastmeanshiftclusteringmethod,wecanperformvideosegmentationatinteractiverate.

Wede neafeaturespace{xi}nsionsforthevideo.Thefeaturevectori=1ofsevendimen-ateachpixelxi=(σppi,σcci,σtti,σ i)comprisesitspositionpi(xandyco-ordinate),colorci(Labcolorvector),timetiandmotion i,thesefourkindsoffeaturescanbeweightedbyparametersσp,σv,σtandσ ,andthevaluesoftheseparametersarede- nedasconstantsforallpixels.AsillustratedinFigure5,thereare1.44×107pixelsinthevideo,andwe rstclusterthevideowithsamplingfactor16348usingKDtree.Ittakesourmethod16.23secondstoperformthemeanshiftcluster-ingonCPU,and1.2secondsonGPU.Ittakes320.3secondsonCPUusing[PD07

paper1220_final

paper1220_final

paper1220_final

paper1220_final

paper1220_final

paper1220_final

].

Figure5:Videosegmentation.Leftcolumn,inputvideo(600×480×50),fromtoptodown,1thframe,28thframe,50thframe.Rightcolumn,segmentationresults.

4.3.Meshmodelsegmentation

Similartoimageandvideosegmentationinimageprocess-ingandanalysis,surfacesegmentationisoneofthemostimportantoperationsingeometryprocessingandmodeling.Yamauchietal.[YLL 05]adoptedthemeanshiftcluster-ingtomeshpartition,andproducedfeaturesensitivemeshsegmentation.InFigure6,wegivethemeshsegmentationusingtheproposedfastmeanshiftcluttering.Wede neafeaturespace{xi}ni=at1ofsixdimensionsforthemeshmodel.Thefeaturevectorvertexxi=(σppi,σvvi)comprisesitspositionpi(x,yandzcoordinate)andnormalvi(threedi-mensionvector)whichcanbeweightedbyparametersσpandσv.Thevaluesofσpandσvarede nedasglobalcon-stantsforallvertices.

InFigure6,usingthevariantofmeanshiftproceduretothemeshmodel,wereceiveapatch-typesegmentationre-sults,andthesegmentationresultsaresensitivetothesalientsurfacefeatures.Furthermore,weadoptthehierarchicalim-agesegmentationmethod[PD07]tomeshmodelandgener-atethehierarchicalsegmentationresults.Notethatourfastmeanshiftmethodisguaranteedtoproducesegmentationre-sultswhichcatchthemeaningfulcomponents,noadditionalcomputationisneededtocomputethehierarchicalresults.Forameshmodelwith70,994vertices,ittakes0.31sec-ondforourmethodtocomputetheresults.Wealsogivethecomparisonresultswith[YLL 05].AsshowninFigure6,ourapproachgeneratesmoreconvincingresults.4.4.Animatedgeometryobjectsegmentation

Theproposedfastmeanshiftalsocanbeusedtoacceleratetheanimatedobject(geometrysurfacesequences)segmen-

c2010TheAuthor(s)Journalcompilation

c2010TheEurographicsAssociationandBlackwellPublishingLtd.

上一页第6页

你可能喜欢

热门文档

相关文档

  • 3080_Final_2012_answers

    Answers Final Exam CHEM 3080 April 9, 2012, 2 pm-5 pm CB 121 Name ... no additional paper, but you are allowed to use the back of the pages...

  • final paper

    Academic paper Final Paper Literature review ? including the backgrand, previous research limitation, purpose, method, findings and results, futhur research, ...

  • final_draft

    final_draft_其它语言学习_外语学习_教育专区。The Final Draft Reflecting on ...8 ? by 11 inch white paper ? double spaced, size 12 font ? one inch...

  • 120703_UNIQLO PR Plan CN final-14时 (NXPowerLite)

    120703_UNIQLO PR Plan CN final-14时 (NXPowerLite)_销售/营销_经管营销_...paper (release) ● 媒体/政府关系 ● 准备并进行面对媒体的准备 ● 紧急时...

  • final-exam paper

    Appendix Final exame paper of systemic anatomy for foreign studnts in WZMC Part One: 5.5 Years MBBS Major, 1 Semester of 2006-2007, Systemic Anatomy...

  • ...Multiprocessor_Architecture_white_paper_Final_v1...

    Tegra_Multiprocessor_Architecture_white_paper_Final_v1.1 tegra2 nvidiategra2 nvidia隐藏>> Whitepaper NVIDIA Tegra Multi-processor Architecture V1.0 P age...

  • 邱媛媛_小微_final

    1000-1220(2010)07- An approach for Building Complex Network of Web Service...In this paper, the approach of building complex network of web service is...

  • English_Final_Examination_Paper_A

    English_Final_Examination_Paper_A_英语考试_外语学习_教育专区。山东建筑大学往年期末及答案College English Achievement Test (Band I ) 试卷类型 A Paper One ...

  • take-home paper final

    take-home paper final_英语学习_外语学习_教育专区 暂无评价0人阅读0次下载举报文档 take-home paper final_英语学习_外语学习_教育专区。Yintong Liu ...

  • 西门子面试题 A6_Answers_to_paper_testing_Final

    西门子面试题 A6_Answers_to_paper_testing_Final_IT/计算机_专业资料。A6答案Answer to paper testing of SCM Excellence Graduate Program 2011 I. 翻译 1. ...

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