Unifying regularization and Bayesian estimation methods for enhanced imaging with remotely sensed Data-part II: implementation and performance issues

Abstract

The fused Bayesian-regularization (FBR) method from a companion paper provides a rigorous theoretical formalism for optimal estimation of the power spatial spectrum pattern (SSP) of the wave field scattered from the probing surface as it is required for enhanced radar imaging of the remotely sensed scenes. Being nonlinear and solution-dependent, the optimal FBR method requires extremely complex nonlinear solution-dependent operator inversions and, therefore, cannot be recommended as a numerically realizable estimator of the SSP. Here, we design a family of robust easy-to-implement FBR algorithms, provide the relevant computational recipes, and discuss their performances. We comment on the practical aspects of the robustified FBR estimators, such as numerical implementation and improvement in the output SNR. The advantage in using the proposed robust FBR method is demonstrated through simulations of enhancing the SAR images formed using the conventional matched filtering of the trajectory signal.