Bayesian Despeckling of SAR Image Based on Logarithmic Transform and Markov Random Fields
Conference: EUSAR 2006 - 6th European Conference on Synthetic Aperture Radar
05/16/2006 - 05/18/2006 at Dresden, Germany
Proceedings: EUSAR 2006
Pages: 4Language: englishTyp: PDFPersonal VDE Members are entitled to a 10% discount on this title
Li, Heng-Chao (National Key Laboratory of Microwave Imaging Technology, Institute of Electronics, Chinese Academy of Sciences, and Graduate School of Chinese Academy of Sciences, P. R. China)
Hong, Wen; Wu, Yi-Rong (National Key Laboratory of Microwave Imaging Technology, Institute of Electronics, Chinese Academy of Sciences, P. R. China)
SAR images are inherently affected by multiplicative speckle noise, which will degrade the human interpretation or computer-aided scene analysis. In order to suppress speckle and preserve significant details, a Bayesian despeckling of SAR image based on logarithmic transform and Markov random fields (MRF) is proposed. Firstly, the logarithmic transform is employed to convert the multiplicative speckle to an additive noise. And the estimation of radar cross section (RCS) is obtained by Maximum a Posteriori (MAP) method, which assumes that the probability density function for likelihood is normally distributed, and Gaussian Markov random fields (GMRF) is chosen as the texture prior in virtue of computational convenience. The proposed method has been experimented on simulated images and real SAR images.