高时相星载序贯SAR图像运动目标检测方法  

Moving Target Monitoring Algorithm Based on High-frame-rate SAR Images

作　　者：陈杰[1] 杨威[1] 王亚敏[1] 李春升[1] CHEN Jie;YANG Wei;WANG Yamin;LI Chunsheng(School of Electronic and Information Engineering,Beihang University,Beijing 100191,China)

机构地区：[1]北京航空航天大学电子信息工程学院,北京100191

出　　处：《雷达学报（中英文）》2022年第6期1048-1060,共13页Journal of Radars

基　　金：国家自然科学基金(62271028)。

摘　　要：该文针对低信杂噪比条件下运动目标检测难的现状,提出了高时相星载序贯合成孔径雷达(SAR)图像运动目标检测方法。首先,根据检测机理的不同将现有星载SAR运动目标检测方法分为3类,并进行了对比分析;其次,基于凝视观测模式建模分析了高帧频序贯SAR图像获取方式;在此基础上,将动目标检测等效为未知尺度、未知到达时间的一维瞬态微弱扰动信号检测,并理论分析了沿时间维高帧频序贯SAR图像间动目标幅度扰动的sinc函数形式,背景杂波幅度的缓变和系统噪声幅度的无规则快变状态;再次,为实现目标和杂波、噪声的可分性,基于核函数机理实现了动目标在高维空间的深度关联;最后,通过仿真和真实数据验证了所提方法的有效性,并分析了检测性能。性能分析结果表明在低信杂噪比条件下所提方法检测性能优于传统的恒虚警类方法。To alleviate the difficulty in monitoring a moving target under a low signal-to-clutter-noise ratio,this paper proposes a moving target monitoring method with high frame-rate spaceborne Synthetic Aperture Radar(SAR)images.First,based on the detection mechanism,current spaceborne SAR moving target detection methods are divided into three categories,and a comparative analysis is performed.Second,the acquisition method of a high-frame-rate SAR image sequence is analyzed based on the staring observation mode.Then,the moving target detection is equated to one-dimensional transient weakly perturbed signal detection with unknown scale and arrival time.Next,the sinc-function form of moving target perturbation between highframe-frequency SAR images,slowly changing background clutter,and irregular fast-changing state of system noise are analyzed theoretically.To separate the target,clutter,and noise,the deep correlation of the moving target in high-dimensional space is realized based on the kernel function mechanism.Finally,the effectiveness of the proposed method is verified by simulation experiments and real SAR data,and under a low signal-toclutter-noise ratio,the detection performance of the proposed method is better than the traditional method of constant false alarm rate.

关 键 词：低信杂噪比 星载合成孔径雷达 微弱扰动 高时相 深度关联 

分 类 号：TN957[电子电信—信号与信息处理]