基于內(nèi)容的圖像和視頻檢索技術(shù)研究及其系統(tǒng)實(shí)現(xiàn).pdf

1、湖南師范大學(xué)碩士學(xué)位論文基于內(nèi)容的圖像和視頻檢索技術(shù)研究及其系統(tǒng)實(shí)現(xiàn)姓名：李紅英申請(qǐng)學(xué)位級(jí)別：碩士專業(yè)：計(jì)算數(shù)學(xué)指導(dǎo)教師：滿家巨20060301AbstractWiththedevelopmentofmultimediaandcomputernetworktechnique，thevolumeofdigitalimageandvideoisgoingupfabulouslyIthasagoodsensetoorganize，manage

2、andutilizethedataoftheimageandvideoefficientlywhichmakesthecontentbasedimageandvideetrievalbeonthetopresearchofthemultimediaareaButtherestillarealotofproblemsneededtoberesolvedinthecontent—basedimageandvideetrievalatpres

3、entInthispaper，wedeveloptwoprototypesystemsstartedwiththeaccuracyandtimeofthecontent—basedimageandvideetrievalOneofthemisacontentbasedimageretrieval，theotherisasemanticbasednewsvideetrievalsystemRelevanceFeedbackandSuppo

4、rtVectorMachinelearningareintroducedintothecontendbasedimageretrievalsystemWealsocompareDTW&GSVMresultstoUFM&GSVM’SwhichprovesthatthealgorithmofDTW&GSVMiseffectiveThefirstshotsegmentationandthesecondshotsegmentationusedi

5、nthesemanticbasednewsvideetrievalsystemmaketheshotssegmentationaccurate；then，basedonthis，weintroduceinteractiveofnewsvideoscenesegmentationThesemanticretrievalofnewsvideoisimplementedbytheannotationofscenethroughmetadata

6、annotationThereare5sectionsinthispaperInsection1boththejobwehavedoneandtheresearchstateofcontentbasedimageandvideetrievalaredescribedSection2introducestheknowledgeofpaaemrecognitionandSupportVectorMachineThedesignandreal

7、izationofaregionbasedimageretrievalprototypesystemisprovidedinsection3Section4describesthedesignandrealizationofasemanticbasednewsvideetrievalsystemAndfinally，weconcludeinsection5togetherwithadiscussionoffutureworkKeyWor