A neural network approach to real-time pattern recognition

This paper presents a new neural network approach to real-time pattern reco gnition on a given set of binary (or bipolar) sample patterns. The percepti ve neuron of a binary pattern is defined and constructed as a binary neuron with a neighborhood perceptive field. Letting its hidden units be the resp ective perceptive neurons of the patterns, a three-layer forward neural net work is constructed to recognize these patterns with minimum error probabil ity in a noisy environment. The theoretical and simulation analyses show th at the network is effective for pattern recognition and can be easily imple mented under strict real-time constraints.