Sea Clutter Texture Estimation: Exploiting Decorrelation and Cyclostationarity

International audienceStatistical model of sea clutter is approximated as the product of speckle and texture characterized respectively by a short and long decorrelation times. This paper introduces two approaches for the estimation of such texture. The first approach exploits the physical properties of the correlation time and the energy of texture and speckle. To identify the texture range profile from one pulse radar frequency to another, we combine the AR estimation of sea clutter auto-covariance with the Empirical Orthogonal Functions. The second approach employs a mixed l2-l1 norm minimization criterion to account for the sparse harmonic structure of the texture