No abstract

The noise radiated from ships in the ocean contains inforniation about their machinery, being nornially used f o r detection and idetttiflcation purposes. In this work we use a neural classifier to identib the radiated noise received by a hydrophone that was far from the ship. The classification is perJ5ornied in the frequency donzain using a feedforward neural network, which is trained using the backpropagation algorithm. It is shown that the use of an averaged spectral irlforniation during the production phase improves sign$-cantly the eficiency of the classi@Tr; when it is conipared to a neural class$ier that processes frequency domain data obtained from individual acquisitior4 windows.

No abstract

No abstract

A neural discriminating analysis is used for classifying passive sonar signals. Preprocessed information from the amplitude spectra of the noise radiated from ships is projected onto only a few principal discriminating components for feeding the input nodes of the neural classifier. Envisaging practical applications, in which new incoming classes not known by the time of the training phase have to be detected in the production phase, a method is provided using the identification of outliers to trigger the arriving of a new class. Using experimental data, it is shown that up to 85% of patterns from an untrained class can be identified, without significant decrease on the efficiency of the classifer for classes known beforehand.