一维卷积神经网络用于雷达高分辨率距离像识别

Radar HRRP Target Recognition with One-dimensional CNN

针对人工提取高分辨率距离像(HRRP)优良特征比较困难的问题,研究了基于一维卷积神经网络(CNN)的HRRP识别方法。利用CNN具有分层学习特征的能力,训练CNN自动地从HRRP中学习有用的特征并分类。在仿真实验中描述了网络的相关配置,分析了不同激活函数、不同参数、不同网络结构的识别性能,对比了CNN与其他分类器的识别结果,用可视化特征图直观地说明了CNN通过卷积层能够学习到易于分辨的特征。实验结果表明CNN具有很好的识别性能。

The one-dimensional Convolutional Neural Network（CNN） is applied to High Range ResolutionProfile（HRRP） recognition to avoid the difficulties of extracting good features manually. The CNN canlearn features hierarchically and extract useful features from HRRP and classify targets automatically. Theconfiguration of the CNN is described, and the recognition perfomtance of different activation functions, parameters, network structures and different classifiers are compared by simulations. Visual feature maps demonstrate that CNN can obtain distinguishable features through convolutional layer. The simulation resultsshow that CNN can get features from HRRP effectively and has a good recognition performance.