Target recognition and classification using neural networks

Intelligent systems such as neural networks and fuzzy logic or a combination thereof have been successfully applied to solving problems in the area of pattern recognition. Neural networks are modeled off the biological neurons in the body and their triggered responses to stimuli; while, fuzzy logic uses human linguistic and expert knowledge to make decisions. In many situations, there are multiple sensors making measurements of the environment. The objective is to combine the measurements (data fusion) in a way that provides an output that describes the environment. The development of a mathematical formulation for solving a problem like this is often very difficult. On the other hand, intelligent systems such as neural networks and fuzzy logic have provide good solutions in complex scenarios like this, where it is difficult to develop closed form mathematical solutions to multi-input/multi-output problems. Therefore, in this paper a neural network approach is used to recognize and classify airborne targets. The target recognition and classification problem was broken into two parts. The first part consists of using a neural network to recognize the existence of a target. Then, given the existence of a target, the next step consists of classifying the target type. Two case studies are presented showing the efficacy of using neural networks for target recognition and classification.