基于门控多尺度匹配网络的小样本SAR目标识别  

Few-shot SAR target recognition based on gated multi-scale matching network

作　　者：刘旗 张新禹 刘永祥 LIU Qi;ZHANG Xinyu;LIU Yongxiang(College of Electronic Science and Technology,National University of Defense and Technology,Changsha 410073,China)

机构地区：[1]国防科技大学电子科学学院,湖南长沙410073

出　　处：《系统工程与电子技术》2022年第11期3346-3356,共11页Systems Engineering and Electronics

基　　金：国家自然科学基金(61025006,60872134,61901482,61921001);中国博士后科学基金(2018M633667)资助课题。

摘　　要：为了解决传统合成孔径雷达(synthetic aperture radar, SAR)目标识别方法在小样本条件下泛化能力差、识别准确率低的问题,通过在匹配网络的基础上引入权重门控单元和多尺度特征提取模块,提出了基于门控多尺度匹配网络的小样本SAR目标识别方法。在该方法中,多尺度特征提取模块能够提取匹配网络不同卷积层的多尺度特征,权重门控单元能够根据不同的识别任务赋予特征不同的权重大小,实现根据具体任务选择最具代表性的目标特征,从而以该特征为主导完成目标识别任务。在运动和静止目标获取与识别(the moving and stationary target acquisition and recognition, MSTAR)数据集上对提出的方法进行了验证,实验结果表明,所提方法较其他3种小样本学习方法和两种小样本SAR目标识别方法表现出了一定的优越性,而且所提方法经实验验证在噪声环境下表现出了一定的鲁棒性。In order to solve the poor generalization ability and low recognition accuracy problems of traditional synthetic aperture radar(SAR) target recognition methods in few-shot condition, a novel few-shot SAR target recognition method based on gated multi-scale matching network is proposed, which introduces weight gated unit and multi-scale feature extraction module into matching network. In the proposed method, the multi-scale feature extraction module is used to extract multi-scale features of different convolutional layers in matching network and the weight gated unit is used to weight different multi-scale features according to different recognition tasks. The proposed method achieves the effect of carrying out different recognition tasks mainly based on features of different layers thanks to the weight gated unit. The proposed method is evaluated on the moving and stationary target acquisition and recognition(MSTAR) dataset and achieved promising performance compared with state-of-the-art few-shot learning methods and few-shot SAR target methods. Furthermore, the proposed method shows good robustness in noisy environments.

关 键 词：合成孔径雷达 雷达目标识别 小样本学习 融合目标识别 度量学习 元学习 

分 类 号：TN95[电子电信—信号与信息处理]