multi-level 思想在图形分割的应用的应用



Figure12:Left:SmoothblendingoftheStanfordbunnyandCyberwareIgeamod-els.Right:offsettingoftheStanforddragonmodel;f=0.0075(top)andf= 0.01(bottom);α=0.





75.
Figure14:LinearmorphingoftwoheadmodelsapproximatedbyMPUimplicits,MaxPlanckHeadandHeadofMichelangelo’sDavid.
Figure13:Linearblendingofoctahedronandcube.
7Discussion
Thispaperdescribesanewimplicitsurfacerepresentationbasedonlocalshapefunctions,partitionsofunity,andanoctreehierarchy.StrengthsoftheMulti-LevelPartitionofUnityformulationinclude:
Fastsurfacereconstructionandrendering.Representationofsharpfeatures.Reconstructionfromincompletedata.
Choiceofeitherapproximationorinterpolationofthedataandtheabilitytoadaptivelyvarytheapproximationaccuracy.
GivenapointsetmodelprocessedbytheMPUmethodwithaspec-i edaccuracy,thecomputationaltimeandmemoryusagedependsonthegeometriccomplexityofthemodel:thehigherthegeomet-riccomplexity,thedeepertheoctreeissubdivided.ThisisclearlydemonstratedbyFigure1wherethereconstructionof nefeaturesrequiresadeepersubdivision.
Table1presentsRAMmemoryusageandcomputationaltimemeasurementsforsimultaneousgeneratingandpolygonizingvari-ouspointsetmodels.Notethatourmethodisquitefast.Ourexper-imentswithstate-of-the-artRBF-based3DsurfacereconstructiontechniquessuchasFastRBF[Carretal.2001]andotherssuggestthattheMPUmethodisconsiderablyfasterthantheseothertech-niques.2
Model
NumberofpointsRelativeerrorPeakRAMNumberoftrianglesComp.time
BunnyBunnyscansDragonBuddhaDavid(2mm)34,611362,272433,375543,6254,124,4542.5×101.0×10 38.0×10 4
0.01.0×10 434MB110MB195MB442

MB

810MB91,104219,676819,704648,332



1,296,5220:071:461:396:53

10:33
Table1:Memoryandcomputationaltimemeasurementsforgenera-tion+polygonizationofMPUimplicitsforvariouspointsetmodels.Computationswereperformedona1.6GHzMobilePentium4with1GBRAM,andtimingsarelistedasminutes:seconds.
2ComparingtheresultsofTable1withthoseofTable2in[Carretal.2001]one
can ndthattheMPUmethodis20-30timesfasterthantheFastRBFtechnique[Carretal.2001].
NoticethatprocessingtimefortheBuddhamodelismorethanthreetimeslongerthenthatforthedragonmodelwhichhasasim-ilarsize.ThisisbecausewereconstructtheBuddhabytheMPUinterpolationwhichrequiresadeepersubdivisionandwidersupportforthecorrespondingpartitionofunityfunctions.
Becauseofitslocalnature,theMPUmethodismoresensi-tivetothequalityofinputdata,especiallythe eldofnormals,tocomparewiththeapproximationandinterpolationtechniquesbasedonglobally-supportedRBFs[Carretal.2001;TurkandO’Brien2002].Nevertheless,accordingtoourexperiments,theMPUmethodissuf cientforaccurateshapereconstructionfromawidevarietyofdatasets.Theparametersinourcurrentimple-mentationoftheMPUapproachareadjustedforprocessingtypicaloutputsoflaserscannerdevices.Webelievethattheparametermodi cationsneededfordifferentclassesofinputarefairlyintu-itiveinordertohandlescattereddataoflower(higher)qualityattheexpenseoflower(higher)computationalspeed.
Unlikethecrustapproach[Amentaetal.1998],theMPUmethodisnotsupportedbyrigorousresultsguaranteeingcorrectrecon-structionofinputdatasatisfyingcertainpropertiesdescribedmath-ematically.Itisapricewepayforahighspeedofourmethod.WewouldlikealsotostressherethatourmethodisnotanRBFtechnique.RBFisaglobalapproximation/interpolationmethodbe-causeofitsglobalvariationalnature.Ourmethodisalocalone.Wemakeuseoftwofunctions,thepartition-of-unityweightsandthelocalpiecewisequadraticapproximationfunctions,whichisdiffer-entthanthesinglefunctionusedbyanRBFapproach.Thistwo-functionapproachgivesbene tssuchassharpfeaturereconstruc-tionthat,todate,havenotbeenpossibleusingRBFs.
SmoothnesspropertiesoftheMPUimplicitsaredeterminedbythoseofweightfunctions(3).ChoosingsmootherweightfunctionswillproducesmootherMPUimplicits.
Similartootherimplicitfunctionshaperepresentationschemes,theMPUimplicitsarenotcapableofrepresentingcorrectlysurfaceswithboundaries.
Weseeanumberofopportunitiestoimproveourapproach.Otherestimationofthedistancefunctionmightbebene cial.Thedistancefunctionisaruledsurfacewithsingularities,thereforeus-ingquadraticfunctionstoapproximatethedistancefunctionisnotoptimal.Aricherlibraryoflocalshapeapproximationscouldbegeneratedinordertoreconstructaccuratelymorecomplexsharpfeatures.Finally,theMPUapproachshouldbewellsuitedtoanout-of-coreimplementationduetothelocalnatureoftheweightingfunctions.
Multi-level Partition of Unity Implicits Di e go S a l ume Octob e r 2 3 rd , 2 0 1 3 A u t h o r: O hta ke , et . a l . Over...
Partition of Unity for a Class of Nonlinear Parabolic Equation on Overlapping Non-Matching Grids_专业资料。维普资讯 http://www.cqvip.com NUMERICAL ATHEMATI...
Two-grid partition of unity method for second order elliptic problems 隐藏>> 维普资讯 http://www.cqvip.com Ap1MahMeh一n1E.20,94:2—3 p.t.c.Eg.d...
Multi-level Partition of... 8页 2财富值 辞职后五险一金怎么处理? 11页 ...a partition of unity method to solve large-scale, scattered data problems....
免费 The partition of unity m... 暂无评价 24页 免费 A Method for the Horizon... 暂无评价 20页 免费 Multi-level Partition of... 8页 2财富值喜欢...
Integrated importance measure for multi-state coherent systems of k level_数学_自然科学_专业资料。Journal of Systems Engineering and Electronics Vo1.24,NJour...
Overview of Partition ... 20页 免费 Multi-level Partition ... 8页 1下载...Engng 2003; 00:1–44 PARTITION OF UNITY ENRICHMENT FOR BIMATERIAL INTERFACE...
Multi-level Partition ... 8页 1下载券 A Particle-Partition o... 暂无评价...'Partition of Unity' RBF Networks are Universal Function Approximators y Dept...
Teleportation of Three-Level Multi-partite Entangled State by a Partial Three-Level Bipartite En_专业资料。维普资讯 http://www.cqvip.com Cmmu.ho.hs(eigCi...
The coupling of partition of unity techniques to level set methods is an ...of natural neighbors that was intro- nodal discretizations in multi-dimensions...
热门文档