Maximum a posteriori speckle filtering of synthetic aperture radar images: A sequential approach

In this paper, a sequential speckle filtering method for synthetic aperture radar (SAR) images is proposed. Maximum a posteriori (MAP) estimation is achieved by incorporating a Gamma prior for the underlying signal. By making the shape parameter of the prior distribution adaptive, image features can be well preserved. The effectiveness of the new algorithm is demonstrated with both simulated and real SAR images. Experimental results show that the proposed method not only has comparable speckle reduction performances with classical methods, but also possesses better feature preservation ability. In addition, it runs very fast because of its sequential nature and so is particularly suitable for processing very large images.