基于自適應分塊的視頻圖像壓縮的研究.pdf

1、分類號UDCY855809密級——單位代碼!Q151基于自適應分塊的視頻圖像壓縮的研究安豐玲指導教師梁德群職稱教授學位授予單位大連海事大學申請學位級別碩士學科與專業(yè)信號與信息處理論文完成同期論文答辯日期2006320答辯委員會主席TheResearchofvideoimagecompressionbasedontheseIfadaptivepartitioningAbstractTheimagecompressionisanimport

2、anttaskinthefieldofimageprocessingTheaimoftheimagecompressionistocutdowntheredundantdataoftheimagewithoutdamagingthevalidinformationoftheimageAtpresent，duringthearithmeticregularlyusedinthefieldofimagecompressingarithmet

3、ic，thedistinctcharacteristhatitaveragelysegmentstheimage，andthencarriesoutthetranslationandcodingtoeveryblockButeveryimageisdifferenttoeachotherandeveryonehasthedifferentdistributionaboutsmoothregionandmarginalregionFrom

4、theviewpointofcompression，compressionwithbigdimensiontothebigsmoothregionandwithsmalldimensiontothesmallmarginalregioncanimproveefficiencyandqualityprominentlyHereinthisarticlebringsforwardthecompressivearithmeticofselfa

5、daptiveblockingFirstofall，segmentstheimagewiththemethodofclusteringbasedonforcefunctionanddifferenttabsaregiventothedifferentregionSecondlyconsideredtothetabs，quadtreepartitioningisusedtosegmenttheimageblockwiththesizeof

6、32andthearithmeticendswhenthesizeoftheblockis4Atlast，thesegmentingblockiscodingwithdifferentdimension，includingDCTtransformbasedondigital，quantification，journeycoding，differencecodingHuffmancodingandsegmentinginformation

7、codingThesegmentationmethodadoptedbythisarticleisefficientandcapable；meanwhile，thefastDCTtransformisdesignedanditbooststheefficientofthewholearithmeticBasedonaboveidea，thisarticleputsforwardselfadaptiveestimationtomotion

8、vectorsandmakestheabecedariandiscussionThefoundationisthatmovementissmalltotheadjacentframewiththeexceptionofscenechanging；thestillregionhasfllargeproportion；andthestrongrelativityexistsintheadjacentframeThemethodisgetti

9、ngridofthestillregionsbasedonthediffframeAtthesametime，selfadaptivesegmentationisusedtothediffframeinordertoestimatethemovevectorsFromtheachievementofarithmeticandtheanalysisofresults，comparedtothetraditionalcompressions