超分辨率重建算法研究.pdf

1、重慶大學碩士學位論文超分辨率重建算法研究姓名：牛羽申請學位級別：碩士專業(yè)：信號與信息處理指導教師：朱冰蓮20090526ABSTRACTImagerestorationtechniqueiswidelyusedImagesuperresolutionreconstructionofimageprocessinginrecentyearshasbeenaresearchhotspotinthefieldofimagerestoration

2、nhasaresearchingsignificanceintheoreticallyaswellasabroadapplicationinpracticallySuperresolutionrestorationisaneconomicimageprocessingmethodthatrestoreshighel“resolutionimagescontainingmoredetailsfromanimagesequenceoflow

3、erresolution，throughsuchmethodsasimagefusion，deblurringandde‘noisingImagesuperresolutionreconstructionbreakstherestrictionsofthefirstgenerationofthetraditionalimagerestorationtechnologyAndwiththestrictoftheoryandextensiv

4、eapplication，ithasmadealargenumberofexperts，researchersandscholarstoparticipateinandexploreLotsofforeignresearchinstitutionsalsohavemadeittobetheirimportantdirectionofstudyThispaperhasacomprehensiveanalysisofcurrentdomes

5、ticandintemationalImagesuperresolutionreconstructiontechniqueresearchAimingattherestorationofdegradedimagesequencesuperresolutionreconstruction，imagesequencedegradedprocessandreconstructionprocessofill—posedprobleminsupe

6、rresolutionreconstructionareAnalyzedinthispaperUsethemethodofleastsquaremethod，adaptiveregularizationanddistributioniterationtoreconstructImagesequenceThefirstiterationisfromthesinglelowresolutionimagetotheinterimimageto

7、avoidcomputationalcomplexityofsportsfuzzymatrixTheseconditerationgreatlyreducesthenumberofiterationwithhighlyrelevantimagesAtthesometime，savememoryspacewhattheIterationrequiredandexertstheadvantagesofbothAlsotheintroduct

8、ionofadaptiveregularizationmethodisused，SOthatimagereconstructioncanbebetterTheexperimentalresultsshowthatthismethodiseffective，andthePSNRofthereconstructionimageatpeakandthesubjectivevisioneffectweresignificantlyimprove