| The next generation of infrared imaging seekers, now in development, will allow for the implementation of more sophisticated and smarter tracking algorithms, able to keep a positive lock on a targeted aircraft in the presence of countermeasures such as decoy flares. A part of the target selection algorithm will be able to use a pattern recognition system which employs features extracted from all possible targets detected in the missile's field of view. Probabilistic neural networks are able to reach performances similar to those of optimal Bayesian classifiers by approximating the probability density functions of the features of the samples used in training the network. The probabilistic approach also has the advantage of generating an output which indicates the confidence level that the network has in its answer. The purpose of this study is to evaluate the performances and the integration of such a network for an infrared imaging seeker emulator used by the Defence Research Establishment Valcartier for countermeasure studies. |
