Despeckling of SAR clutter amplitude data using G0-MAP filter
Dheeren Ku Mahapatra, Soumya Shubhra Ray, Lakshi Prosad Roy · 2016
Synthetic aperture radar (SAR) images are inherently affected by a multiplicative noise known as speckle. In this paper, we propose a despeckling filter based on G0-distribution and formulate maximum a posteriori (MAP) estimator for texture amplitude. Here, the texture is modeled using inverse Gamma distribution which portrays areas with varying degree of heterogeneity. We estimate the roughness parameter of G0-distribution using method of log-cumulants (MoLC). The G0-MAP estimation of texture amplitude is performed on multilook synthetic data and single look real clutter data. Simulation results are presented to illustrate that the proposed G0-MAP filter achieves better mean preservation and speckle reduction compared to state-of-the-art filters, and promises wider application.