| [1] |
Band L E.Topographic Partition of WatershedswithDigitalElevationModels[J].犠犪狋犲狉犚犲狊狅狌狉犮犲狊犚犲狊犲犪狉犮犺,1986,22(1):15 24[2] KongYueping,FangLi,JiangYonglin,etal.ANew MethodofExtractingTerrainFeatureLinesbyMorphology[J].犌犲狅犿犪狋犻犮狊犪狀犱犐狀犳狅狉犿犪狋犻狅狀犛犮犻犲狀犮犲狅犳 犠狌犺犪狀犝狀犻狏犲狉狊犻狋狔,2012,37(8):996 999(孔月萍,方莉,江永林,等.提取地形特征线的形态学新 方 法 [J].武 汉 大 学 学 报 · 信 息 科 学 版,2012,37(8):996 999)[3] HuangPeizhi,LiuZehui.ExtractionofRidgeandValleyfromDEMBasedonGradient[J].犌犲狅犿犪狋犻犮狊犪狀犱犐狀犳狅狉犿犪狋犻狅狀 犛犮犻犲狀犮犲狅犳 犠狌犺犪狀 犝狀犻狏犲狉狊犻狋狔,2005,30(5):396 399(黄培之,刘泽慧.基于地形梯度方向的山脊线和山谷线的提取[J].武汉大学学报·信息科学版,2005,30(5):396 399)[4] HuangPeizhi.A New MethodforExtractingTer rainFeatureLinesfrom DigitizedTerrainData[J].犌犲狅犿犪狋犻犮狊犪狀犱犐狀犳狅狉犿犪狋犻狅狀犛犮犻犲狀犮犲狅犳犠狌犺犪狀犝狀犻狏犲狉狊犻狋狔,2001,26(3):247 252(黄培之.提取山脊线和山谷线的一种新方法[J].武汉大学学报·信息科学版,2001,26(3):247 252)[5] Jin Hailiang,KangJianrong,Gaojingxiang.Re searchonthe Algorithm ofExtracting RidgeandValleyLines Using ContourData[J].犌犲狅犿犪狋犻犮狊犪狀犱犐狀犳狅狉犿犪狋犻狅狀 犛犮犻犲狀犮犲狅犳 犠狌犺犪狀 犝狀犻狏犲狉狊犻狋狔,2005,30(9):809 812(靳海亮,康建荣,高井祥.利用等高线数据提取山脊 (谷)线算法研究[J].武汉大学学报·信息科学版,2005,30(9):809 812)[6] FeiLifan.ExperimentsofGroup GeneralizationofContourLinesonTopographic Maps[J].犑狅狌狉狀犪犾狅犳 犠狌犺犪狀犜犲犮犺狀犻犮犪犾犝狀犻狏犲狉狊犻狋狔狅犳犛狌狉狏犲狔犻狀犵犪狀犱犕犪狆狆犻狀犵,1993,18(Suppl):6 22(费立凡.地形图等高线成组综合的试验[J].武汉测绘科技大学学报,1993,18(增刊):6 22)[7] HuangLina,FeiLifan.ExperimentalInvestigationontheThreeDimensionGeneralizationofContourLinesusing3DD P Algorithm[J].犌犲狅犿犪狋犻犮狊犪狀犱犐狀犳狅狉犿犪狋犻狅狀犛犮犻犲狀犮犲狅犳 犠狌犺犪狀犝狀犻狏犲狉狊犻狋狔,2010,35(1):55 58(黄丽娜,费立凡.采用3DD P算法的等高线三维综合实验研究[J].武汉大学学报·信息科学版,2010,35(1):55 58)[8] DakowiczM,GoldC.ExtractingMeaningfulSlopesfrom TerrainContours[J].犐狀狋犲狉狀犪狋犻狅狀犪犾犑狅狌狉狀犪犾狅犳 犆狅犿狆狌狋犪狋犻狅狀犪犾 犌犲狅犿犲狋狉狔 牔 犃狆狆犾犻犮犪狋犻狅狀狊,2003,13(04):339 357[9] BasriR,CostaL,GeigerD,etal.DeterminingtheSimilarityof DeformableShapes[J].犞犻狊犻狅狀 犚犲狊犲犪狉犮犺,1998,38(15):2365 2385[10]BaiX,LateckiLJ,LiuW Y.SkeletonPruningbyContourPartitioningwithDiscreteCurveEvolution[J].犘犪狋狋犲狉狀犃狀犪犾狔狊犻狊犪狀犱 犕犪犮犺犻狀犲犐狀狋犲犾犾犻犵犲狀犮犲,犐犈犈犈犜狉犪狀狊犪犮狋犻狅狀狊狅狀,2007,29(3):449 462500 第40卷第4期 熊汉江等:一种提取山脊山谷线的新方法犃犖犲狑 犕犲狋犺狅犱狋狅犈狓狋狉犪犮狋犜犲狉狉犪犻狀犉犲犪狋狌狉犲犔犻狀犲狊犡犐犗犖犌犎犪狀犼犻犪狀犵1 犔犐犡犻狌犼狌犪狀11 StateKeyLaboratoryofInformationEngineeringinSurveyingMappingandRemoteSensing,WuhanUniversity,Wuhan430079,China犃犫狊狋狉犪犮狋:Inthispaper,anewmethodfortheextractionofridgeandvalleylinesispresented.Ittakesadvantagesofmean shiftandmultifactorialFallingWaterModel.Themethodconsistsofthreeprinci palsteps:1)Bybuildingthetopologyrelationshipbetweencontours,thestartpointsfortracingcanbeabstractedfromthem.2)Basedontherelationshipbetweencontourbendingsandterrainfeaturelines,terrainpointsaredividedintoconvexesandconcaves.Atthesametime,virtualfan shapedareasarebuilt.3)AccordingtoFalling WaterModelandMaximumProbabilityFlows,ridgelinesandvalleylinesareextracted.Theexperimentalresultspresenteddemonstratethatthefeaturepointscanbewellextractedandthenoisesandmatchingerrorscanbereduced.犓犲狔狑狅狉犱狊:MeanShift;RidgeLines;ValleyLines;TerrainFeatureLines;TerrainAnalysis;FallingWaterSimulation犉犻狉狊狋犪狌狋犺狅狉:XIONGHanjiang,PhD,professor,specializesin3DGIS,WebGIS,andcartographicgeneralization.E mail:xionghanjiang@163.com犆狅狉狉犲狊狆狅狀犱犻狀犵犪狌狋犺狅狉:LIXiujuan,postgraduate.E mail:dykllxj@126.com犉狅狌狀犱犪狋犻狅狀狊狌狆狆狅狉狋:TheNationalHighTehnologyResearchandDevelopmentProgramofChina(863Program)Nos.2013AA12A203,2012AA12A405;ShenzhenSpecialDevelopmentofStrategicEmergingIndustries,檪檪檪檪檪檪檪檪檪檪檪檪檪檪檪檪檪檪檪檪檪檪檪檪檪檪檪檪檪檪檪檪檪檪檪檪檪檪檪檪檪檪檪檪檪檪檪No.ZDSY20120618163343353.(上接第495页)犃犛犛犈２犅犪狊犲犱犕狌犾狋犻犛狅狌狉犮犲犜犲狓狋狌狉犲犅犾犲狀犱犻狀犵犕犲狋犺狅犱犳狅狉犔犪狉犵犲犛犮犪犾犲犞犻狉狋狌犪犾犜犲狉狉犪犻狀犛犻犿狌犾犪狋犻狅狀犠犈犐犢狅狀犵1,2 犇犐犖犌犢狌犾犻狀3,4 犌犗犖犌犌狌犻狉狅狀犵1 犇犝犢犐犖犌1 犣犎犗犝犢犪狀51 InstituteofGeospatialInformation,InformationandEngineeringUniversity,Zhengzhou450053,China2 SichuanEngineeringResearchCenterforEmergencyMapping& DisasterReduction,Chengdu610041,China3 FacultyofGeosciencesandEnvironmentalEngineering,SouthwestJiaotongUniversity,Chengdu611756,China4 InstituteofSpaceandEarthInformationScience,TheChineseUniversityofHongKong,HongKong999077,China5 SchoolofResourcesandEnvironment,UniversityofElectronicScienceandTechnologyofChina,Chengdu610000,China犃犫狊狋狉犪犮狋:Thetexturedataplaysanimportantroleinthree dimensionaltopographysimulationforvir tualbattlefieldenvironment.3DgraphicsenginesuchasOpenGLorD3Dprovidesmulti texturemap pingmechanismorshader,whichisabletoachievetheintegrationofmultipletextures.Butthea chievementofmultipletexturesisGPU based,whichwilltakealargeoccupancyofGPUmemoryandbandwidth.Inthinclients,suchmethodwillcausealowerrenderdelayandlowerefficiency.There fore,anovelSSE2basedmulti textureblendingmethodispresentedinthispaper,whichimplementsthemultipletexturesblendingprocessonCPU.Theexperimentprovedthatthismethodcaneffective lyreducetheGPUload,achievetheintegrationofreal timemulti layeredtexture,andimprovetherenderingefficiencyofthree dimensionaltopographysimulationsystem.犓犲狔狑狅狉犱狊:Multi Sourcetextureblending;SSE2;three dimensionterrainvisualization;large scalevirtualterrainenvironment犉犻狉狊狋犪狌狋犺狅狉:WEIYong,PhDcandidate,specializesininformationextractionandvisualization.E mail:whuwy@163.com犆狅狉狉犲狊狆狅狀犱犻狀犵犪狌狋犺狅狉:DingYulin,PhD,associateprofessor.E mail:rainforests@126.com犉狅狌狀犱犪狋犻狅狀狊狌狆狆狅狉狋:TheNationalNaturalScienceFoundationofChina,Nos.41101354,41471332,41201440,OpenResearchFundbySichuanEngineeringResearchCenterforEmergencyMapping& DisasterReduction,No.K2015B014.
|
百度学术
下载:
