摘要: |
近年来,卷积神经网络(Convolutional Neural Network,CNN)在合成孔径雷达(Synthetic Aperture Radar,SAR)图像目标分类中取得了较好的分类结果。CNN结构中,前面若干层由交替的卷积层、池化层堆叠而成,后面若干层为全连接层。全卷积神经网络(All Convolutional Neural Network, A-CNN)是对CNN结构的一种改进,其中池化层和全连接层都用卷积层代替,该结构已在计算机视觉领域被应用。本文将A-CNN应用于SAR图像目标分类中,针对公布的MSTAR数据集,提出了基于A-CNN的SAR图像目标分类方法,并与基于CNN的SAR图像分类方法进行对比。实验结果表明,基于A-CNN的SAR图像目标分类正确率要高于基于CNN的分类正确率。 |
关键词: 卷积神经网络 全卷积神经网络 合成孔径雷达 目标分类 |
DOI: |
分类号:TP75 |
基金项目:国家自然科学基金项目(面上项目,重点项目,重大项目) |
|
SAR Image Target Classification based on All Convolutional Neural Network |
|
|
Abstract: |
Convolutional neural network (CNN) recently achieved good classification results in synthetic aperture radar (SAR) image classification. In the CNN structure, the preceding layers are stacked by alternating convolutional and pooling layers, while the latter layers are fully connected layers. The all convolutional neural network (A-CNN) is an improvement on the CNN structure, where the pooling and fully-connected layers are replaced by convolutional layers. It was applied in the field of computer vision. In this paper, A-CNN was applied to SAR image target classification. For the published MSTAR data set, one method of SAR image target classification based on A-CNN was proposed and compared with CNN-based method. The experimental results show that the accuracy of SAR image classification based on A-CNN is higher than that of CNN. |
Key words: convolutional neural network all convolutional neural network synthetic aperture radar target classification |