基于神经网络的舰船目标识别研究

为了实现对水下目标的识别,在现有特征提取方法的基础上,提出了一种从DEMON谱线谱和DEMON谱连续谱提取的特征方法,并设计了一个基于BP神经网络和多神经网络分类识别器的舰船目标识别系统。通过对实际舰船噪声目标进行识别,识别效果比较满意。这对舰船目标识别的发展具有一定的参考价值。

Researches for Ship Identification Based on Neural Network

In order to identify ships, the paper put forth a technology of extracting features from line spectrum and continuous spectrum of DEMON, then designing a ships identification system based on BP neural network and multiple neural network weighting algorithm. Finally Use this system to identify real ship noise. The result is acceptable when identifying. It is of reference value for the development of ship identification.