您好,欢迎访问三七文档
当前位置:首页 > 商业/管理/HR > 经营企划 > 高斯模糊图像的复原处理与研究
[]20100305[]863(2006AA060103)[](1977),,2001,,,(,434023)(,124010)[]分析了几种经典图像复原算法,在已知图像退化函数的情况下,对高斯模糊图像分别使用逆滤波维纳滤波有约束的最小二乘方滤波算法进行了复原,在这几种算法的参数选取上得到了丰富的经验数据仿真结果表明,在高噪声环境下,维纳滤波抑制噪声的能力较强,有约束最小二乘方滤波复原方法保持细节效果最好[]图像复原;逆滤波;维纳滤波;有约束的最小二乘方滤波[]TP751[]A[]16731409(2010)02N07704,[1],,,,,,,,[2],,,,,3,1图像退化模型,,f(x,y)11g(x,y),,,[3,4]:g(x,y)=f(x,y)*h(x,y)+n(x,y)(1),h(x,y),(PointSpreadFunction,PSF);*,n(x,y)(1)[4]:g=Hf+n(2),g;f;n,g(x,y)N!N,g(x,y)1gN,2gN,,N2!1,gfn,N2!1,HN2!N2PSFf,[2]2逆滤波复原技术,[5]:(2),n,1ff^,Hf^g,n:∀77∀()2010672:JournalofYangtzeUniversity(NatSciEdit)Jun2010,Vol7No2:Sci&Eng#n#2=nTn=#g-Hf^#2=(g-Hf^)T(g-Hf^)(3)(3),f^:L(f^)=#g-Hf^#2(4)Lf^,M=NH-1,:f^=(HTH)-1HTg=H-1(HT)-1HTg=H-1g(5),(5):F^(u,v)=G(u,v)H(u,v)(6)F^(u,v),,:f^(x,y)=F-1[F^(u,v)]=F-1G(u,v)H(u,v)(7)(7),H(u,v)uv,,,:F^(u,v)=F(u,v)+N(u,v)H(u,v)(8)(8),H(u,v)uv,N(u,v)/H(u,v)H(u,v)u,v,N(u,v),(),1/H(u,v),uv,M(u,v),M(u,v):M(u,v)=tH(u,v)∃d1/H(u,v)(9),td1,d3维纳滤波复原技术,,f(x,y)f^(x,y)e2[4],:e2=minE{[f(x,y)-f^(x,y)]2}(10),E;f(x,y);f^(x,y)(3)f^1Q(),#Qf^#2,a,f^:L(f^)=#Qf^#2+a(#g-Hf^#2-#n#2)(11)(4),(s=1/a):f^=[HTH+sQTQ]-1HTg(12)fnRfRnQ,Rf=E{ffT},Rn=E{nnT},QTQ=R-1fRn(12)(s=1):F^(u,v)=1H(u,v)|H(u,v)|2|H(u,v)|2+Sn(u,v)/Sf(u,v)G(u,v)(13),H(u,v);Sn(u,v)/Sf(u,v);|H(u,v)|2=H*(u,v)H(u,v),H*(u,v)H(u,v);Sn(u,v)=|N(u,v)|2;Sf(u,v)=|F(u,v)|2F^(u,v)f^(x,y),,,H(u,v)0,,,KSn(u,v)/Sf(u,v)(13):F^(u,v)=1H(u,v)|H(u,v)|2|H(u,v)|2+KG(u,v)(14)∀78∀()20106,K0,,,K0,,,K,,,;K,,4有约束的最小二乘方滤波复原技术(12)Q,[4],,,f(x,y)(x,y):2fx2+2fy2%4f(x,y)-[f(x+1,y)+f(x-1,y)+f(x,y+1)+f(x,y-1)](15)f(x,y):p(x,y)=0-10-14-10-10:min2fx2+2fy22:#g-Hf^#2=#n2#(16):F^(u,v)=H*(u,v)|H(u,v)|2+s|P(u,v)|2G(u,v)(17),s,;P(u,v)p(x,y)1235试验结果及分析,matlab31231(a)328!1808BMP;1(b),067,0,2;1(c),(9)t09,d007;1(d),(14)K0005;1(e),(17)s10-3,,,∀79∀72::3,220322032(a)1(b)20;2(b),(9)t007,d013;2(c),(14)K0021;2(d),(17)s10-21,,,,2,,,,MSE[6](MeanSquareError)f(x,y)M!N,MSE:MSE=1M!N!M-1x=0!N-1y=0(f(x,y)-f^(x,y))2(18)13MSEMSE21152710366101342020537196911785213MSE,MSE,,,,3MSE,20,3MSE,,6结语32,,,3;,,,,,,,,[][1][D]:,2004[2][D]:,2004[3],,,Lucy[J],2009,25(5):279~280[4]RafaelC,RichardE[M]:,2007[5][D]:,2007[6],[J],2008,44(14):187~189[]∀80∀()20106Abstract:Oil/gaswellcementationinBurmainshallowreservoirswithhighpressuregasandlowtemperatureultrahighhighdensitywasimplemented,onthebasisofgraincompositiontheoryofcloseaccumulationconception,themodelofcloseaccumulationparticlesizedistributionwasestablishedIncreasingsolidphaseparticleinnerperunitvolumeslurryanddecreasingwater/cementratiorealizehighdensityslurryandimproveslurryperformanceWithparticlesizedistributionmode,byusingvarisizegrainedhematiteparticleofweightingagent,thegradingweightingexperimentwasoptimizedWiththeweightingagent,thehighdensitycloseaccumulationslurrysystemperformanceisgoodFieldcementationqualityisgood,whichverifiesthedependabilityofcloseaccumulationmodelSimultaneously,theadditivesystemsuitstohighdensityslurryisoptimized,wecompletedthefirst28g/cm3ultrahighhighdensityslurrysystemisdevelopedindomesticallyByaddingfibre,thehighdensityslurrysystemhasgoodperformanceinleakresistanceandsealingKeywords:closeaccumulation;hematite;ultrahighhighdensityslurry;fibresealing74ResearchonRecognitionofShoeprintsCombiningOutlineandTexturalFeaturesGUANYan(LianyungangNormalCollege,Lianyungang222006)LIYuanjin(ChuzhouUniversity,Chuzhou239000)Abstract:ThepaperproposesamethodofshoeprintretrievalbasedonoutsidecontourandinsidetexturalfeaturescenteringonshapefeaturesofshoeprintsAspectrationandspecialpointsofsubareasareusedtodescribetheoutsidecontourofshoeprints,andFourierdescriptorsandthechaincodeareappliedtorepresenttheinsidetexturalfeaturesofshoeprintsTheexperimentalresultshowsthattheproposedmethodofshoeprintfeaturesisfast,exactandpracticalKeywords:outlinefeature;texturalfeature;shoeprint;imagerecognition77ProcessandResearchonRestorationofGaussianburredImageFUQingqing(YangtzeUniversity,Jingzhou434023)Abstract:Severalclassicalimagerestorationalgorithmsareanalyzed,inthecaseofknownfunctionoftheimagedegradation,theGaussianblurredimagesarerestoredbyusinginversefiltering,wienerfiltering,constrainedleastsquaresfilteringalgorithm,awealthofempiricaldataontheparameterselectionoftheabovealgorithmswasobtainedMatlabsimulationresultsshowthatinhighnoiseenvironment,theWienerfilteringhasthebetterabilityofsuppressingnoise,constrainedleastsquaresfilteringhasthebesteffecttoremaintheimagedetailsKeywords:imagerestoration;inversefiltering;wienerfiltering;constrainedleastsquares85ResearchandRealizationof3DVirtualCampusBasedonVrMapHUANGChangjun,HULimin,ZHOUQingshan,CAOYuanzhi(HunanCityUniversity,Yiyang413000)Abstract:Onthebasisoffullabsorptioninrelevantresearchresults,itputsforwardaB/SframeworkconstructedvirtualcampusofHunanCityUniversityAccordingtoitsactualsituationandtheneed,itdesignedthedatabasesandfunctionalmodulesVRMapIMSisusedtorelease3Dsceneon3DsoftwareplatformVRMap,VRMapSDKisusedinthedevelopment,thesystemcanachievebrowsingofthreedimensionalenvironment,queryingandlocating,measuringthedistanceandotherfunctionsThevirtualcampussystemprovidesascientific,simpleandvi
本文标题:高斯模糊图像的复原处理与研究
链接地址:https://www.777doc.com/doc-4762334 .html