摘要
SAR图像中高动态运动目标散焦严重,导致传统的动目标检测方法难以实现有效探测。为此,本文利用SAR序贯复数图像特点,提出了基于核函数深度关联的动目标检测方法。首先,分析了高帧频序贯SAR图像获取方式和运动目标频谱特性;其次,借鉴光学图像动目标检测思路,引入了核函数理论,并按照是否有运动目标经过,将SAR图像像元分为两类;再次,将复数图像域动目标检测转换为沿一维时序扰动的检测,理论分析了基于核函数深度关联的运动目标和杂噪分离的机理,联合实部数据和虚部数据目标检测结果,可进一步提高运动目标检测性能,并给出了复数图像域检测流程;最后,开展了低信杂噪比条件下高动态目标检测仿真实验,验证了本文所提方法的有效性。
Moving targets is difficult to detect in synthetic aperture radar(SAR),as its highly dynamic moving makes the synthetic aperture radar(SAR)image normally defocused.This paper proposes a kernel correlation method to detect moving targets based on the characteristics of sequential SAR complex images.Firstly,the working mode for high frame-rate sequential SAR images and the frequency spectrum characteristics are analyzed.Then,considering the moving target detection method in the optical image,kernel function theory is introduced.Each pixel in the SAR image is classified into two categories according to whether moving targets are present or not.Next,the problem of target detection in sequential SAR images is transformed into one-dimensional temporal disturbance detection,the mechanism of separation of moving targets and clutter based on kernel correlation is analyzed,and the corresponding detection method is presented.The combination of the real signal and the imaginary signal could further improve the moving target detection performance.Finally,simulation experiments on air target detection are carried out,and the results of the experiment verify the effectiveness of the proposed method.
作者
陈昊
王亚敏
CHEN Hao;WANG Yamin(Beijing Institute of Tracking and Communication Technology,Beijing 100094,China;College of Electronic Information Engineering Beijing University of Aeronautics and Astronautics,Beijing 100191,China)
出处
《电波科学学报》
北大核心
2025年第3期536-545,共10页
Chinese Journal of Radio Science
关键词
运动目标
核函数
SAR
序贯图像
moving target
kernel function
synthetic aperture radar(SAR)
sequential image
