| 引用本文: |
马彦恒1 , 褚丽娜1 , 杨晓亮2 , 侯晓泽1.采用斜距波数子带划分的SAR 图像自聚焦方法[J].陆军工程大学,2022,(3):12-21
[点击复制] |
|
| |
|
|
| 本文已被：浏览 120次 下载 182次 |
| 采用斜距波数子带划分的SAR 图像自聚焦方法 |
| 马彦恒1,褚丽娜1,杨晓亮2,侯晓泽1 |
| (1 . 陆军工程大学 石家庄校区, 河北 石家庄 050003 ; 2. 中国电子科技集团 第五十四研究所, 河北 石家庄 050000) |
| 摘要:小型旋翼无人机载合成孔径雷达在运动中易受大气湍流等因素影响，产生二维空变的运动误差，影响SAR成像质量。本文提出一种基于拟极坐标系的快速因子分解反向投影（QPG-FFBP）成像自聚焦方法。该方法基于斜距波数子带划分，采用基于加权的相位梯度自聚焦模型估计每个子带的方位相位误差，并融合非系统距离单元徙动的相干性估计，完成全图像的运动补偿，解决二维空变误差补偿问题，提升图像的聚焦效果。通过仿真实验以及小型旋翼无人机载实测SAR数据验证了该方法的有效性。 |
| 关键词: 合成孔径雷达 二维运动误差 自聚焦 拟极坐标系快速因子分解反向投影 运动补偿 |
| DOI：10.12018/j .issn.2097-0730.20211125004 |
| 投稿时间：2021-11-25 |
| 基金项目:国防科技基金(2019-JCJQ-JJ-015) ; 军内科研项目(KYSZJQZL2019) 。 |
|
| SAR Autofocusing Method Based on Slant Range Wavenumber Subband Division |
| MA Yanheng1 , CHU Lina1 , YANG Xiaoliang2 , HOU Xiaoze1 |
| (1 . Shijiazhuang Campus,Army Engineering University of PLA,Shijiazhuang 050003 ,China;
2. The 54th ResearchInstitute of China Electronics Technology Group Corporation,Shijiazhuang 050000,China) |
| Abstract: The small rotor unmanned aerial vehicle synthetic aperture radar (SRUAV SAR) is susceptible to atmospheric turbulence and other factors, which leads to 2-dimensional (2-D) space-variant motion errors and affects the quality of SAR imaging. This paper proposed a quasi-polar grid-based fast factorized back-projection (QPG FFBP) imaging autofocusing method. Based on the slant range wavenumber subband division, the azimuthal phase errors (APE) of each subband was estimated by weighed phase gradient autofocus (WPGA), and the non-systematic range cell migration (NsRCM) was estimated according to the coherence of APE and NsRCM. After fusion of the subband NsRCM, the NsRCM compensation of the whole image was completed, and the 2-D space-variant error compensation problem was solved. This method was verified by simulation and real data of a SRUAV SAR. |
| Key words: synthetic apertureradar (SAR) 2-dimensional (2-D) space-variant motion errors autofocusing quasi-polar grid-basedfastfactorized back-projection (QPG FFBP) motion compensation |
|
