基于分形的圖像編碼技術研究.pdf

1、中南大學碩士學位論文基于分形的圖像編碼技術研究姓名：袁鳳璋申請學位級別：碩士專業(yè)：計算機應用技術指導教師：季桂樹20080520ABSTRACTToeffectivelystoreandtransmitthemassiveimagedataIStheimportantproblemwhichtheinformationagefacesnowBecauseofitsnecessityandfeasibilitythestudyonimag

2、eencodinganddecodingalreadybecomesmoreandmoreactiveFractalimagecodingisaverypromisingcompressiontechniqueIthasbeenpaidgreatattentionsinceitspresencebecauseofitsnovelty，highcompressionrata，resolutionindependenceandrapidde

3、codingInthisdissertation，someselfsimilar，fastfractalimageencodinganddecodingschemeareproposed，asfollows：1Amethodthatestimatesimage’Sselfsimilarityispresentedandthefactordenotingcharacterofimage’Sselfsimilarityisdefinedth

4、roughstudyingtheselfsimilaritymodelofimageExperimentalresultsindicatethatthebiggerselfsimilarityparameters，thebeRerimagecodingefficient2Becauseexistingfastencodingmethodsareoftenatthecostofreconstructedimagequality，orCal

5、lonlyobtainpoorspeed—upratio，thispaperfurtherproposedsomemethodswhichimprovedbothqualityandspeedoffractalencoding：throughcombiningtheselfsimilaritycoefficientandclusteringalgorithm，usingtheaverageofrangeblocksvarianceast

6、hecenterofinitializedclusteringandsortingselfsimilarityparameters，itreducingsearchtime，optimizingthealgorithmintheprocessofaffinetransformationwithreducingthematchingdomainsExperimentindicatesthemethodwhichthispaperpropo

7、sedacceleratefractalimageencoding3ThemethodmakesthebestuseofthetheoremofresolutionindependentinfractaltheoryItimprovesaffinetransformationdecodingalgorithmandcombinesaffinetransformationanditeratingtheaverage‘ofrangebloc