1、Subject：ComputerAp衛(wèi)1icationTechn0109Y一●__■■____●____J■_●■___●■■_____■●■__■_●__■●■_■■■■■_●__■■■__●■■●_●■_●■■_____●_●__o___■_____●■__■_●■■_______●■___●______■■●■_一C011ege：gQ!里叟g殳Q至gQ里乜旦主金!魚i里旦Q金一SubmittedDate：|I壘旦墜墾!Y星Q：圣Q

2、!!浙江大學(xué)碩上學(xué)位論文AbstractAbstractTWodimensionalcartoonisVeD，popularnowadaysforitsrichexpressiVenessandspecials夠leItisalsowidelyappliedinmanyareas，such嬲adVertisement，education，aIldentertainmentSofar，ala唱eamountoftwodimensional

3、canoonshaVebeencreatedThecreationoftwo—dimensionalcartoonisabomcomplexa11dtime—consumingjobEa’ectiVereuseofexistingcartoonsc鋤reduceworkloadofaIlimators，mereforeimproVingmeemciencyofcartooncreationThjsp印erfocusesonhowteus

4、echaractermotionintv舊dimensionalcartoon，aIldaccordinglyproposes趾interactiVeca喲onmotioncaptu】rea11dreta唱etingteclllliqueThisteclllliquemainlyconsistsofthreestepsFirst，acharacterposesequenceisextractedfomgiVencanoonsequenc

5、eWeproposeaIlextractionmethodbysupportVectormachines(SVM)Thismethodconvertscharacterposeextractionintoat、Ⅳo—classclassificationproblembySVMSecond，thedefo咖ationofeVer)rpanofthecharacteruIldergoesinthemotionisc印turedfromth

6、eextractedposesequenceWbproposeakey行amebaSedcartoonmotioncapturemethodWitlltllecontoursmarkedbyuserinkeyfames，thecontoursofeVe巧partininlIe姍ediatef鋤escanbelocatedautomaticallyBasedonthecontoursequenceofeaChpart，thedeforma

7、tiondataofeachpartc趴beobtainedFinally，tlleobtainedpandefomationis印pliedtoanewcanooncharactert0generateanewcartoonsequenceFirst，tllepancorrespondencebe觚eenthenewcharacter鋤dtheoriginalcharacterisbuiltT11en，theobtaineddef0咖

8、ationdataiSappliedt0tIlecorrespondingpartsoftheIlewcharacterTheexperimentsdemonstratethatt11eproposedinteractiVecartoonmotionc邳咖IreaIldretargetingtecllIliquec鋤generateast)，lesimilarmotionofanewcharacterf如meXistingcharact