Abstract: This article proposes to make use of the inherent characteristics of synthetic aperture radar (SAR) images to improve the Gibbs-Markov random field (MRF) model for recovering the SAR images. Further, it segments a SAR image into the target and the shadow by means of the theory of connectivity in digital morphology. The new method uses not only Gamma distribution to replace the traditional Rayleigh distribution in the estimate of a maximum a posteriori probability (MAP), but also the connectivity model of pixel intensity value relevance to better extract the goal in the neighborhood of the SAR image pixel space. This method takes full advantage of the relevance between the information of digital morphology of the SAR image and the pixel intensity, and is able to eliminate isolated points and obtain good segment results. Simulation tests illustrate the validity of the method.

Key words: Markov processes, Gamma distribution, synthetic aperture radar, image recovery, image segmentation, connected clustering