欢迎访问中国科学院大学学报,今天是

中国科学院大学学报

• • 上一篇    下一篇

基于对比度最优和低阶轨道多项式估计的星载SAR自聚焦方法*

陈颖1,2,3, 孟大地1,2†, 李光祚1,2, 黄丽佳1,2, 李莹莹4, 辛煜4   

  1. 1 中国科学院空天信息创新研究院,北京,100094;
    2 中国科学院空间信息处理与应用系统技术重点实验室,北京,100190;
    3 中国科学院大学电子电气与通信工程学院,北京,100049;
    4 北京市遥感信息研究所,北京,100192
  • 收稿日期:2023-02-13 修回日期:2023-04-11 发布日期:2023-04-11
  • 通讯作者: †E-mail: mengdd000151@aircas.ac.cn
  • 基金资助:
    *中国科学院青年创新促进会支持项目(No.2019127)

A novel approach for autofocus in spaceborne SAR based on low-order polynomial orbit model and contrast maximization

CHEN Ying1,2,3, MENG Dadi1,2†, LI Guangzuo1,2, HUANG Lijia1,2, LI Yingying4, XIN Yu4   

  1. 1 Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100094, China;
    2 Key Laboratory of Technology in Geo-spatial Information Processing and Application System, Chinese Academy of Sciences, Beijing 100190, China;
    3 School of Electronic, Electrical and Communication Engineering, University of Chinese Academy of Sciences, Beijing 100049, China;
    4 Beijing Institute of Remote Sensing Information, Beijing 100192, China
  • Received:2023-02-13 Revised:2023-04-11 Published:2023-04-11

摘要: 高分辨率星载合成孔径雷达(SAR)成像要求精确的卫星轨道位置信息,而实际中GPS卫星可能无法满足精度要求,进而导致成像质量下降。在对散焦星载SAR图像进行自聚焦处理时,我们发现卫星轨道的平滑特性对于处理非常有利。基于该特性,本文提出了一种新的星载SAR自聚焦方法。首先将卫星轨道在三维空间分别用低阶多项式进行拟合,从而将自聚焦问题转化为对多项式的估计问题;然后,选取图像中分散的几个小图像切片,通过最大对比度法得到每个切片中心点的最优距离历程;再根据这几个距离历程调整整景轨道,最终得到能使各切片良好聚焦的轨道估计结果。通过合适的切片选取,进而可以使整幅图像良好聚焦。仿真实验和实际数据结果验证了该方法的有效性。

关键词: 星载SAR, 自聚焦, 轨道误差, 对比度最大

Abstract: Spaceborne SAR imagery with higher spatial resolutions requires greater knowledge of the satellite's orbit. However, sometimes GPS systems are not available to provide accurate position information that the image resolution requires. In our experience, the smoothness of spaceborne SAR orbit can provide help for autofocus. In this paper, a novel approach for autofocusing in spaceborne SAR is proposed. First, the modeling of the orbit in three-dimensional as polynomial functions is involved. Therefore, autofocusing can be achieved by estimating the polynomial coefficients. Then, several patches distributed over the SAR image are selected, and the optimal range history of the center point in each patch is obtained based on the maximum-contrast optimization. The estimated orbit of the whole scene can be derived through the range history information. Finally, the image can be refined with better focusing performance. The estimated orbit is capable of satisfying the optimal quality for every patch. Furthermore, with proper patch selection, including the number and the relative location of patches, better quality within the whole scene can be reconstructed by the estimated orbit. This method is tested and validated with simulation experiments and real data.

Key words: spaceborne synthetic aperture radar (SAR), autofocus, orbit error, contrast maximization

中图分类号: