Small military and industrial under-surface objects detection using data from the UWB GPR, MLP filter, and oscillatory neural network

Abstract

In the paper, we present a solution to the problem of recognizing subsurface objects detected by UWB radar using a hybrid neural network consisting of an MLP filter, a Hilbert block, and an oscillatory neural network. Based on information resonance, this hybrid neural network architecture recognizes low-amplitude noisy signals that are signs of subsurface objects. In addition, the proposed neural network model can filter reflected signals with interference and amplify valuable signals due to resonance and has higher performance than existing models of artificial neural networks.