Adaptive restoration of speckled SAR images

The paper proposes a restoration method for speckled images generated by coherent imaging systems (e.g., synthetic aperture radar synthetic aperture sonar, ultrasound imaging, and laser imaging). These systems are invariably affected by speckle noise and therefore restoration/filtering of the mean backscattered signal (backscattering coefficient) is often necessary. The approach is Bayesian: the observed image is assumed to be a realization of a random field built upon the physical mechanism of image generation; the backscattering coefficient image is modelled by a compound Gauss-Markov random field which enforces smoothness on homogeneous regions while preserving discontinuities between neighboring regions. The maximum a posteriori probability (MAP) criterion is adopted. An expectation maximization type iterative scheme embedded in a continuation algorithm is used to compute the MAP solution. Application examples performed on radar real and synthetic data are presented.