Low complexity DCT-based distributed source coding with Gray code for hyperspectral images

在线阅读 下载PDF 导出详情
摘要 Tocompresshyperspectralimages,alowcomplexitydiscretecosinetransform(DCT)-baseddistributedsourcecoding(DSC)schemewithGraycodeisproposed.UnlikemostoftheexistingDSCschemes,whichutilizetransforminspatialdomain,theproposedalgorithmappliestransforminspectraldomain.Set-partitioning-basedapproachisappliedtoreorganizeDCTcoefficientsintowaveletliketreestructureandextractthesign,refinement,andsignificancebitplanes.TheextractedrefinementbitsareGrayencoded.Becauseofthedependencyalongthelinedimensionofhyperspectralimages,lowdensityparitycheck-(LDPC)-basedSlepian-WolfcoderisadoptedtoimplementtheDSCstrategy.Experimentalresultsonairbornevisible/infraredimagingspectrometer(AVIRIS)datasetshowthattheproposedparadigmachievesupto6dBimprovementoverDSC-basedcoderswhichapplytransforminspatialdomain,withsignificantlyreducedcomputationalcomplexityandmemorystorage.
机构地区 不详
出版日期 2010年06月16日(中国期刊网平台首次上网日期,不代表论文的发表时间)
  • 相关文献