Target recognition based on directional filter banks and higher-order neural networks

A new approach for the classification of SAR targets is presented here, whi ch combines maximally decimated directional filter banks with higher-order neural networks (HONNs). HONNs are neural networks that can achieve perform ance similar to that of standard multilayered neural networks, but without the hidden layer. Their performance can be made invariant to geometric tran sformations of the input imagery in the design process, while their computa tional complexity can be reduced by employing a preprocessor to reduce the dimensionality, such as coarse coding. Most past image classifiers using HO NNs have been designed for carefully thresholded binary images. However, ge nerating useful binary representations that; can be used as inputs can be d ifficult for modalities such as SAR. As an alternative, we use a novel HONN implementation that accepts gray-level input pixels using directional filt er banks. In order to do this, a new modified tree-structured directional f ilter bank structure in a very computationally efficient form is introduced . The performance of the proposed approach is demonstrated and compared in imagery taken from the public MSTAR database. The new approach is shown to be effective in enhancing the discrimination power of the HONN inputs, lead ing to significantly improved performance. (C) 2000 Academic Press.