子波神经网络及其在自动目标识别中的应用

Application of WNN in Automatic Target Recognitno

本文分析了于波神经网络（ＷＮＮ）的生理依据，构造了便于启发式特征抽取的子波。研究了该子波及其相应的ＷＮＮ在非相干雷达船目标特征抽取中的应用。对现场实录的四类船目标回波数据进行实验，结果表明新构造的低维特征具有较高的目标分类应用价值，抗噪性能良好．

This paper concerns the application of wavelet neural network(WNN)in feature ex-traction for incoherent radar ship target recognition. The biological of WNN has been analysed.And, a new kind of analysis wavelet, which is suitable for heuristic feature extraction, has been con-structed.Experiments are carried out for the proposed method with four kinds of practical radar ship target video-echo data and the results indicate that the new feature vector in low dimension is valua-ble for target classificatlon,and the noise-resistance character is good.