视音频信息融合算法分析-analysis of video and audio information fusion algorithm.docxVIP

  • 3
  • 0
  • 约4.01万字
  • 约 59页
  • 2018-08-01 发布于上海
  • 举报

视音频信息融合算法分析-analysis of video and audio information fusion algorithm.docx

视音频信息融合算法分析-analysis of video and audio information fusion algorithm

太原理工大学硕士研究生学位论文THEFUSIONALGORITHMRESEARCHOFVIDEOANDAUDIOINFORMATIONABSTRACT10recentyears,withthedeveJopmentofthecomputersinformationization,moreandmorevideoequipmentsaswellasthetechnologyappJiedinpeoplesdailylife.Videoconference,videosearchenginetechnologyaswellasvideοdataioquirytechnologyandsoon.whichwereincludingthemassivenon-textdatum企omthesourceoftelevision,themeetinnotes,thescientificliteratureandsoon.Totheone,thepopularityofpersonalphotographicequipment,andtheimprovementoflnternettechnology,sothatordinarypeopJepublishpersonalvideorecordinghasbecomeextremelysimple,andhaveconsequentlyalargearnountofvideodata.1tisaseveretestfortheexistingvideoprocessingtechnologythathowtodealwithsomanymulti-mediainformation,howtoorganizedataandmaketheindexforittoretrieve.Earlymultimediainformationretrievalalgorithmhasdeviatedfromtheoriginalpurposeofeasyoperations,且阳reretrievalalgorithmsneedtointegratelow-levelfeaturewhichcontainsmorerepresentativefeatureofthevision,hearing,semantic.Themulti-modalprop创yofvideoinformationprovideabasisforinformationfusion.VideoThemostexistinganalysistoolsforthesinglemodalofvideofeatures,butthevideoisthespecialdatawhichhasmulti-modalna阳reandtheresarelevanceaboutthemulti-modalofthevideofeatureswhenitdescribesthesarnetopics.Thereforeneedakindofeffectivemethodtofusionvideofeaturesformoreaccuratelyclassi乌,andretrieval.Them句orworkat出eprocessofvideofeatureandthefusionprocessofvideofeaturesareasfollows:IIIThecuπentmodelofvideodataprocessinglimitedtheapplicationofnews,advertisingandotherspecificareas,andthetechnologyusedinprocessingissingleandold.Tbisarticleusesavideoprocessingtechnologywhichisrelativeefficiencytodefinearelativelycompletemodelofv?deoretrievalpretreatment.lnthismodel,usedthemulti-modalna阳reofthevideolow-levelfeaturetoextractthevideotimestructure,tofeatureextractionforthecontent,andconstructedthesubsetofvideodatafromtheoriginalvideo.Basedonthisprocesstoextractthekeyframesofvideoandextractaudiofeaturesformvideosaudiostreams.Forsimplifythecalcul

您可能关注的文档

文档评论(0)

1亿VIP精品文档

相关文档