SAR Image Denoising Based on Orthogonal Bandlet Domain with Context Modeling
This paper proposes a novel method on SyntheticAperture Radar(SAR )Image denoising,which isbased on context modeling combined with multiscaleorthogonal bandlet coefficients.For the main influenceof SAR image is multiplicative speckle noise,thelogarithm uniform transform is used to convert it toadditive noise.In this paper,the multiscale orthogonalbandlet transform is applied to noisy SAR Image.Thecontextual model is used to compute contextual valuesof bandlet coefficients on each scale.Therefore,bandlet coefficients are reconstructed into some partsby contextual values.The spatially adaptive soft-threshold(T)is to denoise SAR image in each part.Experimental results demonstrate that this methodoutperforms the methods of Ehance-Lee(E-lee),contextual model based on wavelet domain(Wavelet-context),and Hidden Markov Tree model bases onwavelet domain(Wavelet-bin0 on there indexes:Floating index(FI),Standard deviation and Edgesaving index of vertical direction,by using a real SARimage.However,the method can not only reduce thespeckle noise but also preserve edges and highfrequency information,and gets considerable visualeffect on SAR Image denoising.
Anfa Zhang Jingwen Yan Zhixi Wang
Dept.of Electronic Engineering,Shantou University
国际会议
厦门
英文
1179-1184
2008-11-17（万方平台首次上网日期，不代表论文的发表时间）