The fused Bayesian maximum entropy-variational analysis method for computer reconstruction of remote sensing imagery

We address the aggregated Bayesian maximum entropy-variational analysis (BMEVA)-based algorithm for high resolution radar image enhancement and denoising. The use of the variational analysis (VA) approach is formalized by imposing the metrics structures in the corresponding signal spaces. A new formalism for combining the Bayesian maximum entropy (BME) strategy with the VA paradigm is developed. The advantages in image enhancement and denoising achieved using the proposed BMEVA method is illustrated through numerical simulations.