基于小波分解和改進(jìn)型PCNN的圖像融合方法研究.pdf

1、中南大學(xué)碩士學(xué)位論文基于小波分解和改進(jìn)型PCNN的圖像融合方法研究姓名：胡藝齡申請學(xué)位級別：碩士專業(yè)：計算機(jī)科學(xué)與技術(shù)指導(dǎo)教師：鄒北驥20100501ABSTRACTWiththerevolutionofnewtechnologyOurworldentryintothefullinformationeraAnewtheoryandtechniquewasdevelopedwhichisreallyeffectiveinthemulti—

2、sourceinformationfusionFusionasavisualinformationfusion，hasbecomeawidespreadconcernformostpeoplearoundtheworldInourarticle，wecombineswavelettransformwithpulsecoupledneuralnetworksinimagefusionThemainresearchcontentsinclu

3、dethefollowingthreepartsQWecomparedanddiscussedsometraditionalpixellevelimagefusionalgorithmsjustlikesimpleimagefusion，thefusionmethodbasedonpyramiddecompositionandthefusionmethodbasedonwavelettransformThewidelyusedwavel

4、ettransformisdescribedindetailWealsoexplainedhowtousethewaveletdecompositionforimagefusionprocess，andspecificallydescribesthefastwavelettransformalgorithmComparedwithtraditionalwavelettransform，thefastwavelettransformhav

5、ethebettertime—frequencycharacteristics(g)Pulsecoupledneuralnetworkmodelwasintroducedinthepaperweanalysistheparametersinthemodel，andfindintraditionalmethodstheconnectioncoefficientisafixedvalueInordertogetridofthelimitat

6、ion，wepresentusingimagespatialfrequencyafterwavelettransformtodynamicallyadjusttheconnectioncoefficient(亙)Aimingatthelimitationoftraditionalwavelettransform，ourpaperproposeanewimagefusionnovelthatafterdecomposingtheimage

7、withfastwavelettransform，weusepulsecoupledneuralnetworkstochoosefusionruleinsteadofartificialselectionThekeypointisinallusiontothefeatureofhighfrequencyandlowfrequencyusingdifferentrules，respectivelyTheoptiononhi曲frequen

8、cypartistousepulsecoupledneuralnetworksAscarryinglinleinformationofimage，andeffectinglittleonthefinalfusionOnthelowfrequencypart，westillchoosethetraditionalweightedaveragemethodThroughperformingexperimenttotwosetsofimage