Statistical modeling of sea clutter in high-resolution SAR images using generalized gamma distribution

Statistical modeling of sea clutter in synthetic aperture radar (SAR) imagery is fundamental for SAR image interpretation. In this paper, we adopt a recently proposed generalized gamma distribution (GΓD) for modeling sea clutter in high-resolution SAR images. Based on parameter decoupling, an estimator of GΓD, named as scale-independent scale estimation (SISE), is derived, which only refers to several basic operations and can be easily realized. Modeling experiments are carried out over the L-band polarimetric SAR images acquired by JPL/AIRSAR and a VV-polarized C-band TerraSAR-X SAR image. Experimental results show that the advantage of GΓD for modeling sea clutter in high-resolution SAR images is evident comparing to the classic distributions of sea clutter in SAR images including the Weibull, Log-normal and K distributions.