预览加载中，请您耐心等待几秒...

基于参数估计的降晰函数辨识及图像复原算法(英文).docx

预览
1/2
2/2

在线预览结束，喜欢就下载吧，查找使用更方便

下载提示 文本预览

如果您无法下载资料，请参考说明：

1、部分资料下载需要金币，请确保您的账户上有足够的金币

2、已购买过的文档，再次下载不重复扣费

3、资料包下载后请先用软件解压，在使用对应软件打开

基于参数估计的降晰函数辨识及图像复原算法(英文) Introduction Imagerestorationhasbecomeanessentialtaskinvariousfieldssuchasmedicalimaging,remotesensing,digitalphotography,andcomputervision.Thequalityofanimagecanbedegradedduetonoise,distortion,orblurcausedbyvariousfactorssuchassensornoise,atmosphericturbulence,ormotionblur.Inrecentyears,severalapproacheshavebeenproposedtorestorethedegradedimage.Oneofthevitalapproachesisbasedontheparametricestimationofdegradationfunctions. Inthispaper,wepresentastudyontheparametricestimationofdegradationfunctionsandthedevelopmentofanimagerestorationalgorithm.Essentially,thedegradationfunctiondescribestherelationshipbetweentheobservedimageandtheoriginalimage.Ourprimaryfocusistoestimatetheparametersofthedegradationfunctionusingstatisticalmethodsand,basedontheestimatedparameters,torestorethedegradedimage. ParametricEstimationofDegradationFunctions Parametricestimationisastatisticaltechniqueusedtoestimatetheparametersofafunctionbasedonobserveddata.Inimagerestoration,thedegradationfunctionisamathematicalmodelthatexplainsthedegradationprocessofanimage.Themodelcanbedefinedintermsofitsparameters,whichcanbeestimatedusingaparametricestimationmethod.Themostcommonlyusedparametricestimationmethodsarethemaximumlikelihoodestimation(MLE)andtheleastsquaresestimation(LSE). TheMLEmethodisastatisticalapproachusedtoestimatetheparametersofthemodelthatmaximizethelikelihoodfunction.Thelikelihoodfunctionisdefinedastheprobabilityofobservingthedatagiventhemodel.TheMLEmethodiswidelyusedinimagerestorationbecauseitprovidesestimatesthatareconsistentandefficientundercertainconditions.TheLSEmethodisanotherstatisticalapproachusedtoestimatetheparametersofthemodelthatminimizethesumofthesquarederrorsbetweentheobserveddataandtheestimateddata.TheLSEmethodiswidelyusedinlinearregressionanalysisandisrelativelysimpletoimplement. Degradationfunctionscanbeclassifiedintotwocategories:pointspreadfunctions(PSFs)anddegradationoperators.PSFsarefunctionsthatdescribehowasinglepointintheoriginalimageistransfor