Hopfield and pulse coupled neural networks aggregation for enhanced change detection in remote sensing imagery

This paper presents a new technique for high resolution enhancement of Remote Sensing (RS) imagery degraded in a random propagation channel and contaminated with composite noise (additive and multiplicative). The proposed method aggregates two Neural Network (NN) paradigms: The Modified Hopfield Neural Network (MHNN) and The Pulse Coupled Neural Network (PCNN). In the fused strategy, we propose the MHNN technique with the objective to provide the enhanced RS image reconstruction followed by the PCNN algorithm that performs precise change detection. We apply the PCNN for the target detection, segmentation and classification in the reconstructed RS image. Computer simulations examples are reported to illustrate the usefulness of the aggregated unified PCNN-MHNN technique for enhance change detection.