Radar target imaging and identification using the advanced inverse scattering technique

Abstract

A new target recognition procedure is developed. In order to efficiently obtain feature vectors for target discrimination, the closed-form expression of geometrical wave fronts is also derived to provide efficient and accurate computation. Then, the resulting low dimensional feature vectors, obtained by the developed extractor, are identified using the well-known artificial neural networks (ANNs) classifier. In the illustrative experiments. three thin-wire targets are discriminated. The results show that the presented scheme gives successful automatic target recognition (ATR) in the low SNR with low computational costs. Therefore, the proposed technique has a significant potential for use in ATR.