图像高斯噪声及椒盐噪声去噪算法分析-计算机应用技术专业论文.docx

复口.人学她丨：学位论义关键词：图像处理、图像去噪、高斯噪声、椒盐噪声、稀疏编码、多项式拟合、变分滤波、边缘保持势函数、卜丨适应屮值滤波中图分类号：TN 9112U 口.人学她I:学位论文A bstra ctImagedenoisingisahotresearchtopicinthefieldofdigitalimageprocessing,andalotofmodelsandalgorithmsofdenoisinghavebeenproposed.Inhumanhistory，atestimonytotheireraimagedenoisinghassuchagreatinfluenceonmodernsociety,scienceandtechnology.Withthedevelopmentofcomputertechnology,imagedenoising problem hasbeenattractedmoreandmoreattention.D enoisingmethods aredividedintothefollowingthreecategories:methodsbasedonprobabilityandstatisticsmodel,methods basedonvariationalandpartialdifferentialequationsandmethodsbasedontheFourierandwavelettransform.Accordingtotherelationshipbetweenthe imageand thenoise,thenoisemodelismainlydivided into additivenoiseandmultiplicativenoise.Inthispaper,methodsbasedonprobabilityandstatisticsm odeland thevariationalwereusedtorem ovetheadditivenoiseinim age.Thefirstchapterintroducesthebackgroundandclassificationofnoisemodelforresearch inthefieldofimagedenoising.Atthesametime,thesubjectiveandobjectivecriteriaofevaluationofthedenoisingisgiven.Briefintroductiontothefieldofdenoisingmethodsandtheorieswasgiven,aswellasthenoiseclassificationcategoriesbasedondifferentstandards.Inthispaper,researchwejustpayattentiontoGaussiannoiseandsaltandpeppernoise.Thesecondchapterdescribesthesparsecodingim agedenoisingalgorithmbasedonpolynomialfitting.Sparsecodingtheoryanditsapplicationwasgiven.Abriefreviewofthesparsecodingalgorithmforimagedenoisingwasgiven,andwepointedoutitsshortcom ings.Thesim ulationresultsshowtheeffectivenessoftheproposedalgorithm.Thethirdchapterdescribesthedenoisingalgorithm basedonpartialdifferentialequationsandthevariationalmethod.Themethodusesacontinuousdomainofthinkingtodealwithdiscretedigitalimage.First,establishtheenergyfunctiontodenoisingtheimage.Second,achievethepurposeofdenoisingbyminimizing theenergyfunction.Sim ulationresultsshowthatthemethodhasgoodperformanceintheremovaloflow-densitySaltandPepperNoise.Howeverwhenthedensityofimagenoiseincre