Syntactic Pattern Recognition for Radar Target Identification.
O.S. Sands, Frederick D. Garber · 1987
Abstract : A syntactic pattern recognition system is proposed for use in radar target identification. The system utilizes one of three different level-crossing based pattern representation schemes. Likelihood functions are estimated from relative frequency densities and are used for classification of the symbolic pattern representation. Grammatical inference is used to derive a syntax analysis algorithm from the likelihood function classifier performance is estimated using Monte-Carlo simulations, Directions of further research on this promising topic are discussed.