Electromagnetic target classification using time-frequency analysis and neural networks

This paper demonstrates the feasibility and advantages of using a self-orga nizing map (SOM)-type neural network classifier for electromagnetic target recognition. The classifier is supported by a novel feature extraction unit in which the Wigner distribution (WD), a time-frequency representation, is utilized for the extraction of natural-resonance-related energy feature ve ctors from scattered fields. The proposed target classification technique i s tested for a set of canonical targets, displaying an excellent performanc e in terms of both real-time classification speed and accuracy, even in the presence of noisy data. (C) 1999 John Wiley & Sons, Inc.