Targets Detection Based on Neural Network under Background of Sea Clutter
Xiaojun Wen · Jisuanji fangzhen · 2007
To detect echo wave from targets under the background of sea clutter, a new method based on neural network was brought up. Radial based function network can be trained to predict the chaotic dynamic characters of sea clutter. By analyzing the power spectrum of predicting error, the target detection could be made and its speed and distant information. were extracted The emulation results on the conditions of shore-based and ship-borne radar with practical parameters prove to be better than perceptron threshold detection and two-dimension FFT for targets’ information. The relative error of detection is controlled within 10-3~10-4 magnitude. Then the lowest signal-to-noise ratios on each condition were given.