基于.三维模型的自动语义标注的研究与应用

《基于.三维模型的自动语义标注的研究与应用》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。

1、基于三维模型的自动语义标注研究与应用按照上述的思想和算法，搭建了基于三维模型场景语义标注的原型系统。系统通过将新建的三维模型转换为二维图像，与数据库中的样本模型进行图形比较找出异同点，从而为新模型自动生成语义标注。综上所述，本文设计了一套基于三维模型和视频资源的自动语义标注系统框架。本文所提供的三维模型以及视频资源的自动语义标注方案实现了海量三维模型和视频的有效管理。通过对该标注系统的设计方案进行实验验证后，证明此设计方案能够满足对应需求。关键词：三维模型，语义标注信息，WebGL,自动语义标注，视频资源,图

2、像识别。II万方数据基于三维模型的自动语义标注研究与应用3Dmodel-basedSemanticAnnotationManagementResearchandApplicationABSTRACTInrecentyears,alongwiththedevelopmentofthree-dimensionaltechnology,3Dgraphiccomputinghasprovenitsimportantpositioninmanymajorfields.Asthe3D-relatedresourcesha

3、sbeenincreasedrecently,3Dsceneandobjectmodelinghavegainedalotofattractionsandsupportsfromvariouscommunities.Unfortunatelysoftwareproductsthatarecapableofmanipulatingtheseworksarestilllimitedinmanagementaspect,makingitdifficulttoextendthemtotheirfullestpoten

4、tial.Usersnowadaysdemandhigherrequirementsfromthelower-layermanagecomponentsofa3Dsoftware.First,thesoftwareproductsmustbeabletoprovideapowerfulsearchengine,whichcanfilterthroughalargeamountof3Dobjectstofindthedesired3Dmodelsorscenesfromprovidedinformation.S

5、econd,thestructure,characteristicortypeofa3Dobjectmustbeanalyzedinareasonablespeedtosupportthetraditionalsemanticsearch.Theserequirementsleadtoacreativeideaofannotatingthesemanticcontentsto3Dobjects.Semanticannotationhelpusersnotonlymaintainingtheinternalda

6、taandstructure,butalsothesemanticcontentsofa3Dobject.Thesesemanticcontentsincludethecolor,characteristic,typeandotherinformationsuchastheusageforapplications.Accordingtotheabovediscussedproblemsandsolutions,thispaperdesignthesystemcanbefullyautomatedsemanti

7、cannotationprocessforthree-dimensionalmodel.Systemtomeettheusertoquicklyfindthethree-dimensionalmodelfromthedatabasecorrespondingsemanticannotationinformationimported.Orsystemfilescanbeautomaticallyconvertedintoathree-dimensionalmodelofatwo-dimensionalimage

8、.Tosupportsemanticinformationqueryandsearch.Orquicklyidentifythecharacteristicsandclassificationmodelforthree-dimensionalmodelTofulfillthefeatureofcombiningallthesemanticcontentstogether,theannotatingp