Unified experiment design, Bayesian minimum risk and convex projection regularization method for enhanced remote sensing imaging

Abstract

We address new approach for enhanced multi-sensor imaging in uncertain remote sensing (RS) operational scenarios. Our approach is based on incorporating the projections onto convex solution sets (POCS) into the descriptive experiment design regularization (DEDR) and fused Bayesian regularization (FBR) methods to enhance the robustness and convergence of the overall unified DEDR/FBR-POCS procedure for enhanced RS imaging. Computer simulation examples are reported to illustrate the efficiency and improved operational performances of the proposed unified DEDR/FBR-POCS imaging techniques in the extremely uncertain RS operational scenarios. © 2009 Springer-Verlag Berlin Heidelberg.