| 摘要: |
| 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。 |
| 关键词: 目标识别 高分辨率距离像 卷积神经网络 |
| DOI： |
|
| 基金项目:国家自然科学基金资助项目（61401252） |
|
| Radar HRRP target recognition with one-dimensional CNN |
| YIN Heyi,GUO Zunhua |
| (School of Mechanical,Electrical & Information Engineering,Shandong University,Weihai,Weihai 264209,China) |
| Abstract: |
| The one-dimensional Convolutional Neural Network(CNN) is applied to High Range Resolution Profile(HRRP) recognition to avoid the difficulties of extracting good features manually.The CNN can learn features hierarchically and extract useful features from HRRP and classify targets automatically.The configuration of the CNN is described,and the recognition performance 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 results show that CNN can get features from HRRP effectively and has a good recognition performance. |
| Key words: target recognition high range resolution profile convolutional neural network |
《电讯技术》编辑部 