A Neural Network Recognition Model Based on Ship Acoustic-Magnetic Field

Abstract

Underwater recognition system uses targets' passing characteristic curves to recognize ship target. Formerly extraction target characteristic was mainly aimed at one certain physical field, but little analysis was done to the correlation between fields. In order to improve the probability of underwater recognition, this passage makes analysis of the correlation of ship acoustic filed and magnetic field, brings into the concept of ship M-S diagram, then makes research on using neural network technique to recognize ship pattern and designs a neural network ship recognition model. Finally, experimental data is used to testify this model. The result shows that it has high recognition probability on different kinds of ships.