解决扩充现实中的遮挡问题.docx

  1. 1、本文档共6页,可阅读全部内容。
  2. 2、原创力文档(book118)网站文档一经付费(服务费),不意味着购买了该文档的版权,仅供个人/单位学习、研究之用,不得用于商业用途,未经授权,严禁复制、发行、汇编、翻译或者网络传播等,侵权必究。
  3. 3、本站所有内容均由合作方或网友上传,本站不对文档的完整性、权威性及其观点立场正确性做任何保证或承诺!文档内容仅供研究参考,付费前请自行鉴别。如您付费,意味着您自己接受本站规则且自行承担风险,本站不退款、不进行额外附加服务;查看《如何避免下载的几个坑》。如果您已付费下载过本站文档,您可以点击 这里二次下载
  4. 4、如文档侵犯商业秘密、侵犯著作权、侵犯人身权等,请点击“版权申诉”(推荐),也可以打举报电话:400-050-0827(电话支持时间:9:00-18:30)。
查看更多
解决扩充现实中的遮挡问题

第30卷第3期2001年6月电子科技大学学报JournalofUESTofChinaVol.30No.3Jun.2001ResolvingOcclusionProbleminAugmentedRealityShenJieWuYonghongLiuHaowu(CollegeofHydroelectricEng.,SichuanUniversityChengdu610065)AbstractAnewapproachforresolvingocclusionprobleminaugmentedrealityispresentedinthispaper.Theheartofthisalgorithmistotrackatleastfourfiducialpointsfromframetoframethatconstructtheaffineframe,thustheocclusioncontoursbetweenvirtualandrealobjectscanbereconstructedandreprojected.Further,eachcontourislabeledasbeing“behind”or“infrontof”inthefirsttwoframes,dependingonwhetheritisinfrontoforbehindthevirtualobject,sotheocclusionisresolved.Twotypicalexamplesshowthatthealgorithmisfeasibleforresolvingocclusion.Keywordsocclusion;augmentedreality;affineinvariants;contours;reconstructionAugmentedreality(AR)isatechnologybywhichuser?sviewoftherealworldisaugmentedwithadditionalinformationfromacomputermodel,thusmoreinformationspacesareshown[1].Generally,suchsystemsonlyblendvirtualimagerywithrealimagesandattempttominimizeregistrationerrors.However,thismethodisnoteffectivewhenocclusionsexistbetweenvirtualandrealobjects.Thisocclusionproblemcouldeasilybesolvedonconditionthatthemodelof3Dsceneisgiven.Sincelittleisusuallyknownabouttherealworldtobeaugmented,itbecomeschallengingtoresolveocclusioninaugmentedreality.Theoretically,adensemapmaybeinferredfromastereopair,sothedepthbetweenvirtualandrealobjectscanbecompared[2,3].Asamatteroffact,thismethodlacksaccuracy,andisdifficulttouse.Inthispaper,wepresentacontour-basedapproachwith3Dreconstruction.Asitisknown,thereareseverallevelsofreconstruction,includingtheprojective,affine,andEuclideanreconstruction.Differentvisiontasksrequiredifferentlevelsofreconstruction.Onlyanaffinereconstructionisrequiredforfixationpointtracking[4].First,fourormorefiducialpointsfromframetoframewhichconstructanaffinebasisaretracked.Second,thekeypointsofoccludingcontoursbetweenvirtualandrealobjectsmaybespecifiedinteractivelyaccordingtoepipolarandotherconstraintsinthefirsttwoframes.Finally,since

文档评论(0)

pangzilva + 关注
实名认证
内容提供者

该用户很懒,什么也没介绍

1亿VIP精品文档

相关文档