Radar target identification based on feature extraction performed with RBF artificial neural networks

An artificial neural network (ANN) approach for radar image processing is presented in this paper. A renewal concept of simple adaptive units as a foundation for network assembling allows one to design ANN-based feature extraction scheme for 2D-signal processing. It was shown that ANN implementing radial basis function (RBF) processing units can be applied for identification of radar targets described by the set of scatterers. The obtained results indicate a high accuracy estimation of separate scatters centers.