何松华1,张润民1,欧建平2,张军2.基于卷积神经网络的高分辨率雷达目标识别[J].湖南大学学报:自然科学版,2019,(8):141~148
基于卷积神经网络的高分辨率雷达目标识别
High Resolution Radar Target Recognition Based on Convolution Neural Network
  
DOI:
中文关键词:  高分辨距离像  雷达目标识别  卷积神经网络  批归一化  支持向量机
英文关键词:High Resolution Range Profile(HRRP)  radar target recognition  Convolution Neural Network (CNN)  Batch Normalization(BN)  Support Vector Machine(SVM)
基金项目:
作者单位
何松华1,张润民1,欧建平2,张军2 (1.湖南大学 信息科学与工程学院湖南 长沙 410082 2.国防科技大学 ATR实验室湖南 长沙 410073) 
摘要点击次数: 23
全文下载次数: 18
中文摘要:
      提出一种基于卷积神经网络(Convolution Neural Network,CNN)的高分辨率雷达目标识别方法.首先针对小样本应用于深度CNN时训练过程中损失函数值收敛速度慢的问题,利用结合批归一化算法的改进CNN网络对高分辨距离像(High Resolution Range Profile,HRRP)进行自动特征提取;再利用支持向量机(Support Vector Machine,SVM)对距离像特征进行分类.使用军事车辆高保真电磁仿真数据对提出的方法进行验证,识别结果证明了该方法的有效性.
英文摘要:
      A new method of high resolution radar target recognition based on Convolution Neural Network (CNN) was presented. To solve the problem of slow convergence of loss function values during the training process when small samples are applied to the deep CNN, High Resolution Range Profile (HRRP) features were firstly extracted by using the improved CNN combined with the Batch Normalization (BN) algorithm, and then classified by using a Support Vector Machine (SVM). The experimental results using high-fidelity electromagnetic simulation data of military vehicles validate the effectiveness of the proposed method.
查看全文  查看/发表评论  下载PDF阅读器
关闭