Near Real-Time Enhancement of Fractional SAR Imagery Via Adaptive Maximum Entropy Neural Network Computing

Abstract:
We address a neural network (NN) computing-based approach to the problem of near real-time enhancement of compressed fractional SAR imagery. The proposed approach employs the recently developed descriptive experiment design regularization (DEDR) framework for multimode image reconstruction/fusion aggregated with the variational analysis (VA) image enhancement paradigm implemented computationally via the new speeded-up adaptive maximum entropy neural network (MENN) processing technique. The developed DEDRVA- optimal MENN enhancement technique outperforms the recently proposed competing methods both in the achievable resolution enhancement and the convergence rates that is verified via reported computer simulations.