Mathematics and Visualization SeriesEditors GeraldFarin Hans-ChristianHege DavidHoffman ChristopherR.Johnson KonradPolthier MartinRumpf Georges-Pierre Bonneau Thomas Ertl Gregory M. Nielson Editors Scientific Visualization: The Visual Extraction of Knowledge from Data With 228 Figures ABC Georges-PierreBonneau GregoryM.Nielson UniversiteGrenobleI DepartmentofComputerScienceandEngineering Lab.LMC-IMAG IraA.FultonSchoolofEngineering BP53,38041GrenobleCX9 ArizonaStateUniversity France Tempe,AZ85287-8809 E-mail:[email protected] USA E-mail:[email protected] ThomasErtl UniversityofStuttgart VisualizationandInteractiveSystems Institute(VIS) Universitätßtraße38 70569Stuttgart Germany E-mail:[email protected] LibraryofCongressControlNumber:2005932239 MathematicsSubjectClassification:68-XX,68Uxx,68U05,65-XX,65Dxx,65D18 ISBN-10 3-540-26066-8SpringerBerlinHeidelbergNewYork ISBN-13 978-3-540-26066-0SpringerBerlinHeidelbergNewYork Thisworkissubjecttocopyright.Allrightsarereserved,whetherthewholeorpartofthematerialis concerned,specificallytherightsoftranslation,reprinting,reuseofillustrations,recitation,broadcasting, reproductiononmicrofilmorinanyotherway,andstorageindatabanks.Duplicationofthispublication orpartsthereofispermittedonlyundertheprovisionsoftheGermanCopyrightLawofSeptember9, 1965,initscurrentversion,andpermissionforusemustalwaysbeobtainedfromSpringer.Violationsare liableforprosecutionundertheGermanCopyrightLaw. SpringerisapartofSpringerScience+BusinessMedia springeronline.com (cid:1)c Springer-VerlagBerlinHeidelberg2006 PrintedinTheNetherlands Theuseofgeneraldescriptivenames,registerednames,trademarks,etc.inthispublicationdoesnotimply, evenintheabsenceofaspecificstatement,thatsuchnamesareexemptfromtherelevantprotectivelaws andregulationsandthereforefreeforgeneraluse. Typesetting:bytheauthorsandTechBooksusingaSpringerLATEXmacropackage Coverdesign:design&productionGmbH,Heidelberg Printedonacid-freepaper SPIN:11430032 46/TechBooks 543210 Preface ScientificVisualizationisconcernedwithtechniquesthatallowscientistsandengi- neers to extract knowledge from the results of simulations and computations. Ad- vancesinscientificcomputationareallowingmathematicalmodelsandsimulations tobecomeincreasinglycomplexanddetailed.Thisresultsinacloserapproximation torealitythusenhancingthepossibilityofacquiringnewknowledgeandunderstand- ing.Tremendouslylargecollectionsofnumericalvalues,whichcontainagreatdeal of information, are being produced and collected. The problem is to convey all of thisinformationtothescientistsothateffectiveusecanbemadeofthehumancre- ativeandanalyticcapabilities.Thisrequiresamethodofcommunicationwithahigh bandwidthandaneffectiveinterface.Computergeneratedimagesandhumanvision mediated by the principles of perceptual psychology are the means used in scien- tific visualization to achieve this communication. The foundation material for the techniquesofScientificVisualizationarederivedfrommanyareasincluding,forex- ample, computer graphics, image processing, computer vision, perceptual psychol- ogy, applied mathematics, computer aided design, signal processing and numerical analysis. This book is based on selected lectures given by leading experts in Scientific VisualizationduringaworkshopheldatSchlossDagstuhl,Germany.Topicsinclude user issues in visualization, large data visualization, unstructured mesh processing forvisualization,volumetricvisualization,flowvisualization,medicalvisualization andvisualizationsystems.ThemethodsofvisualizingdatadevelopedbyScientific Visualizationresearcherspresentedinthisbookarehavingbroadimpactontheway other scientists, engineers and practitioners are processing and understanding their datafromsensors,simulationsandmathematicsmodels. Wewouldliketoexpressourwarmestthankstotheauthorsandrefereesfortheir hardwork.WewouldalsoliketothankFabienVivodtzevforhishelpinadminister- ingthereviewingandeditingprocess. Grenoble, Georges-PierreBonneau January2005 ThomasErtl GregoryM.Nielson Contents PartI MeshesforVisualization AdaptiveContouringwithQuadraticTetrahedra BenjaminF.Gregorski,DavidF.Wiley,HenryR.Childs,BerndHamann, KennethI.Joy .................................................... 3 OntheConvexificationofUnstructuredGrids fromaScientificVisualizationPerspective Joa˜oL.D.Comba,JosephS.B.Mitchell,Cla´udioT.Silva ................. 17 BrainMappingUsingTopologyGraphsObtained bySurfaceSegmentation FabienVivodtzev,LarsLinsen, BerndHamann,KennethI.Joy,BrunoA.Olshausen...................... 35 ComputingandDisplayingIntermolecularNegativeVolumeforDocking ChangHaLee,AmitabhVarshney .................................... 49 OptimizedBoundingPolyhedra forGPU-BasedDistanceTransform RonaldPeikert,ChristianSigg ....................................... 65 Generating,Representing andQueryingLevel-Of-DetailTetrahedralMeshes LeilaDeFloriani,EmanueleDanovaro................................ 79 Split’NFit:AdaptiveFitting ofScatteredPointCloudData GregoryM.Nielson,HansHagen,KunLee,AdamHuang ................. 97 VIII Contents PartII VolumeVisualizationandMedicalVisualization RayCastingwithProgrammableGraphicsHardware ManfredWeiler,MartinKraus,StefanGuthe,ThomasErtl,WolfgangStraßer..115 VolumeExplorationMadeEasyUsingFeatureMaps KlausMueller,SarangLakare,ArieKaufman...........................131 FantasticVoyageoftheVirtualColon ArieKaufman,SarangLakare .......................................149 VolumeDenoisingforVisualizingRefraction DavidRodgman,MinChen .........................................163 EmphasizingIsosurfaceEmbeddings inDirectVolumeRendering ShigeoTakahashi,YurikoTakeshima,IsseiFujishiro,GregoryM.Nielson ....185 DiagnosticRelevantVisualization ofVascularStructures ArminKanitsar,DominikFleischmann,RainerWegenkittl,MeisterEduard Gro¨ller .........................................................207 PartIII VectorFieldVisualization CliffordConvolutionandPatternMatching onIrregularGrids JuliaEbling,GerikScheuermann.....................................231 FastandRobustExtraction ofSeparationLineFeatures XavierTricoche,ChristophGarth,GerikScheuermann ...................249 FastVortexAxisCalculationUsingVortexFeatures andIdentificationAlgorithms MarkusRu¨tten,Hans-GeorgPagendarm ...............................265 TopologicalFeaturesinVectorFields ThomasWischgoll,JoergMeyer......................................287 PartIV VisualizationSystems GeneralizingFocus+ContextVisualization HelwigHauser ...................................................305 Contents IX Rule-basedMorphingTechniques forInteractiveClothingCatalogs AchimEbert,IngoGinkel,HansHagen ................................329 APracticalSystemforConstrainedInteractiveWalkthroughs ofArbitrarilyComplexScenes LiningYang,RogerCrawfis .........................................345 ComponentBasedVisualisation ofDIETApplications RolfHendrikvanLengen,PaulMarrow,ThiesBa¨hr,HansHagen,Erwin Bonsma,CefnHoile ...............................................367 FacilitatingtheVisualAnalysis ofLarge-ScaleUnsteadyComputationalFluidDynamicsSimulations KellyGaither,DavidS.Ebert........................................385 EvolvingDataflowVisualizationEnvironments toGridComputing KenBrodlie,SallyMason,MartinThompson,MarkWalkleyandJasonWood ..395 EarthquakeVisualizationUsingLarge-scaleGroundMotion andStructuralResponseSimulations JoergMeyer,ThomasWischgoll......................................409 AuthorIndex ....................................................433 Part I Meshes for Visualization Adaptive Contouring with Quadratic Tetrahedra BenjaminF.Gregorski1,DavidF.Wiley1,HenryR.Childs2,BerndHamann1,and KennethI.Joy1 1 InstituteForDataAnalysisandVisualization UniversityofCalifornia,Davis bfgregorski,dfwiley,bhamann,[email protected] 2 BDivisionLawrenceLivermoreNationalLaboratory [email protected] Summary. We present an algorithm for adaptively extracting and rendering isosurfaces of scalar-valued volume datasets represented by quadratic tetrahedra. Hierarchical tetra- hedral meshes created by longest-edge bisection are used to construct a multiresolution C0-continuousrepresentationusingquadraticbasisfunctions.Anewalgorithmallowsusto contourhigher-ordervolumeelementsefficiently. 1 Introduction Isosurface extraction is a fundamental algorithm for visualizing volume datasets. Most research concerning isosurface extraction has focused on improving the per- formance and quality of the extracted isosurface. Hierarchical data structures, such asthosepresentedin[2,10,22],canquicklydeterminewhichregionsofthedataset containtheisosurface,minimizingthenumberofcellsexamined.Thesealgorithms extract the isosurface from the highest resolution mesh. Adaptive refinement algo- rithms[4,5,7]progressivelyextractisosurfacesfromlowerresolutionvolumes,and controlthequalityoftheisosurfaceusinguserspecifiedparameters. Anisosurfaceistypicallyrepresentedasapiecewiselinearsurface.Fordatasets thatcontainsmooth,steepramps,alargenumberoflinearelementsisoftenneeded to accurately reconstruct the dataset unless extra information is known about the data. Recent research has addressed these problems with linear elements by using higher-ordermethodsthatincorporateadditionalinformationintotheisosurfaceex- tractionalgorithm.In[9],anextendedmarchingcubesalgorithm,basedongradient information, is used to extract contours from distance volumes that contain sharp features.Cellsthatcontainfeaturesarecontouredbyinsertingnewverticesthatmin- imizeanerrorfunction.Higher-orderdistancefieldsarealsodescribedin[12].This approachconstructsadistancefieldrepresentationwhereeachvoxelhasacomplete descriptionofallsurfaceregionsthatcontributetothelocaldistancefield.Usingthis representation,sharpfeaturesanddiscontinuitiesareaccuratelyrepresentedastheir exact locations are recorded. Ju et al. [11] describe a dual contouring scheme for 4 B.F.Gregorskietal. adaptivelyrefinedvolumesrepresentedwithHermitedatathatdoesnothavetotest forsharpfeatures.Theiralgorithmusesanewrepresentationforquadricerrorfunc- tions to quickly and accurately position vertices within cells according to gradient information.Wileyetal.[19,20]usequadraticelementsforhierarchicalapproxima- tionandvisualizationofimageandvolumedata.Theyshowthatquadraticelements, insteadoflinearelements,canbeeffectivelyusedtoapproximatetwoandthreedi- mensionalfunctions. Higher-orderelements,suchasquadratictetrahedraandquadratichexahedra,are used in finite element solutions to reduce the number of elements and improve the qualityofnumerical solutions [18].Since few algorithmsdirectlyvisualize higher- orderelements,theyareusuallytessellatedbyseverallinearelements.Conventional visualization methods, such as contouring, ray casting, and slicing, are applied to theselinearelements.Usinglinearelementsincreasesthenumberofprimitives,i.e. trianglesorvoxels,thatneedtobeprocessed.Methodsforvisualizinghigher-order elementsdirectlyaredesirable. We use a tetrahedral mesh, constructed by longest-edge bisection as presented in[5],tocreateamultiresolutiondatarepresentation.Thelineartetrahedralelements usedinpreviousmethodsarereplacedwithquadratictetrahedra.Theresultingmesh definesaC0-continuous,piecewisequadraticapproximationoftheoriginaldataset. Thisquadraticrepresentationiscomputedinapreprocessingstepbyapproximating thedatavaluesalongeachedgeofatetrahedronwithaquadraticfunctionthatinter- polatestheendpointvalues.Aquadratictetrahedronisconstructedfromthecurves alongitssixedges.Atruntime,thehierarchicalapproximationistraversedtoapprox- imatetheoriginaldatasettowithinauserdefinederrortolerance.Theisosurfaceis extracteddirectlyfromthequadratictetrahedra. The remainder of our paper is structured as follows: Section 2 reviews related work. Section 3describes what quadratic tetrahedra are, and Sect. 4describes how theyareusedtobuildamultiresolutionrepresentationofavolumedataset.Sections5 describeshowaquadratictetiscontoured.OurresultsareshowninSect.6. 2 PreviousWork Tetrahedral meshes constructed by longest-edge bisection have been used in many visualization applications due to their simple, elegant, and crack-preventing adap- tiverefinementproperties.In[5],fine-to-coarseandcoarse-to-finemeshrefinement is used to adaptively extract isosurfaces from volume datasets. Gerstner and Pa- jarola[7]presentanalgorithmforpreservingthetopologyofanextractedisosurface usingacoarse-to-finerefinementschemeassuminglinearinterpolationwithinatetra- hedron.Theiralgorithmcanbeusedtoextracttopology-preservingisosurfacesorto perform controlled topology simplification. In [6], Gerstner shows how to render multipletransparentisosurfacesusingthesetetrahedralmeshes,andin[8],Gerstner andRumpfparallelizetheisosurfaceextractionbyassigningportionsofthebinary tree created by the tetrahedral refinement to different processors. Roxborough and Nielson [16] describe a method for adaptively modeling 3D ultrasound data. They