AN APPLICATION OF NEURAL NETWORKS IN AUTOMATIC TARGET RECOGNITION AT SEA

An approache to sea target recognition is examined in this paper. The automatic recognition of the target of observation is considered as detection of its tonnage class, approximate size and resulting maneuvering abilities. A neural network method is used to solve this problem. A number of experiments on real targets are described. Possibilities for application of Automatic Target Recognition (ATR) in sea navigation are proposed based on these experimental findings.