摘要
提出了联合目标区域和阴影的合成孔径雷达(Synthetic Aperture Radar,SAR)目标识别方法。该方法采用椭圆傅里叶描述子描述目标区域和阴影的边界。根据目标区域和阴影的相对位置和大小关系,定义了描述其相对关系的特征矢量。采用稀疏表示分类器对目标区域和阴影的傅里叶描述子以及相对关系矢量分别进行分类,分类的结果利用决策层的线性加权方法进行科学融合。基于融合后的相似度判断目标类别,实现稳健的目标识别。采用MSTAR (Moving and Stationary Target Acquisition and Recognition)公共数据集进行了目标识别实验,验证了方法的有效性。
This paper proposes a Synthetic Aperture Radar( SAR) target recognition method based on joint use of target region and shadow region. Elliptical Fourier Descriptors( EFDs) are used to describe the contours of the target region and shadow region. And a feature vector is designed to describe the relationship between the target region and shadow region. Then,the sparse representation-based classification( SRC) is employed to classify EFDs of the target region and shadow region as well as the feature vector.Finally,a decision-level fusion is conducted to fuse the three results for robust target recognition. To validate the effeteness of the proposed method,experiments are conducted on public MSTAR( Moving and Stationary Target Acquisition and Recognition) dataset.
作者
夏朋举
XIA Peng-ju(Xuchang Vocational Technical College,Henan Xuchang 461000,China)
出处
《中国电子科学研究院学报》
北大核心
2019年第10期1062-1067,1087,共7页
Journal of China Academy of Electronics and Information Technology
关键词
合成孔径雷达
目标识别
目标区域
阴影
稀疏表示
决策层融合
Synthetic Aperture Radar(SAR)
target recognition
target region
shadow region
spare representation-based classification
decision-level fusion
作者简介
夏朋举(1982—),男,河南人,硕士,讲师,主要研究方向为计算机技术,图像处理分析。E-mail:18008265@qq.com。
