Chinese Journal of Computers   Full Text
Title	SAR Image Denoising Based on Lifting Directionlet Domain Gaussian Scale Mixtures Model
Authors	BAI Jing HOU Biao WANG Shuang JIAO Li-Cheng
Address	(Institute of Intelligence Information Processing, Xidian University,Xi’an 710071) (Key Laboratory of Intelligent Perception and Image Understanding of Ministry of Education, Xidian University, Xi’an 710071)
Year	2008
Issue	No.7(1234—1241)
Abstract &   Background	Abstract In this paper, a new speckle suppression method for SAR image is proposed. By combining Directionlet transform with a version of the hidden Markov model—Gaussian scale mixtures (GSM), the marginal distributions of neighbor coefficients in the lifting Directionlet domain are modeled. For removing the speckle noise, the Bayes least square estimation is adopted to evaluate each coefficient. Being regarded as a novel multiscale geometrical analysis tool, Directionlet transform retains the separable filtering, computation simplicity and filter design from the standard two-dimensional wavelet transform, which can capture anisotropic geometrical structures efficiently by multi-direction selection. The introduction of lifting scheme reduces computation amount greatly. Neighborhoods of coefficients at adjacent positions and scales are modeled as the product of two independent random variables: A Gaussian vector and a hidden positive scalar multiplier. Under this model, the marginal of neighbor coefficients are well described and the strong correlation among the amplitudes of neighbor coefficients is also presented adequately. Experiments using plentiful real SAR images indicate that the proposed method outperforms the spatial filters and other methods based on wavelets in terms of speckle reduction as well as image detail preservation. Keywords SAR image; Directionlet transform; Gaussian scale mixtures(GSM); lifting scheme; speckle noise Background In spite of the success of the standard wavelet transform in image processing in recent years, the efficiency of its representation is limited by the spatial isotropy of its basis functions built in the horizontal and vertical directions. Sparse representation of geometrical features in images, as very important visual perception elements, has been a topic of high interest in the last decade. Many multiscale geometric analysis (MGA) methods have been proposed and applied successfully in image processing and analysis. Directionlets is regarded as a new MGA tool which provides a much wider framework for an efficient description for multi-dimension signals. The transform remains the separable filtering and subsampling and the simplicity of computations and filter design from the standard two-dimension wavelet transform. During the research on SAR image processing, this paper provides a denoising algorithm based on lifting Directionlet domain GSM model, which achieves a better denoised result compared with the wavelets and some other conventional approaches. This work is supported by the National High Technology Research and Development Program (863 Program) of China under grant No.2007AA12Z136; the National Basic Research Program of China (973 Program) under grant No.2006CB705700; the National Natural Science Foundation of China under grant No.60672126; the National Research Foundation for the Doctoral Program of Higher Education of China No.20050701013. This group has been working on the mechanisms and applications of MGA tools, including image processing (e.g. image denoising, image fusion, image segmentation and etc). So far, they have obtained some achievements in this area. Lots of papers are published in the international proceedings and journals on this topic.