联合误差估计的机载超高分辨率SAR成像  

Very high resolution SAR imaging method combined with motion estimation

作　　者：景国彬[1,2] 李宁 孙光才[1] 邢孟道[1] JING Guobin;LI Ning;SUN Guangcai;XING Mengdao(National Key Lab. of Radar Signal Processing,Xidian University,Xi an 710071,China;Shanghai Institute of Satellite Engineering,Shanghai 200240,China)

机构地区：[1]西安电子科技大学雷达信号处理国家重点实验室,陕西西安710071 [2]上海卫星工程研究所,上海200240

出　　处：《西安电子科技大学学报》2019年第3期1-7,共7页Journal of Xidian University

基　　金：国家重点研发计划(2017YFC1405600);中央高校基本科研业务费专项资金(JB180213)

摘　　要：针对机载超高分辨率合成孔径雷达存在运动误差严重影响成像质量的问题,提出一种联合误差估计的超高分辨率成像方法。首先,依据合成孔径雷达成像几何,将惯导投影到斜距平面坐标系,反演出运动误差,通过插值运算完成距离空变粗补偿;接着,为了有效结合运动补偿,采用改进施托克插值的距离徙动算法完成徙动校正;利用子孔径误差相位提取和全孔径拼接的方法,完成运动误差的精估计和补偿;采用图像偏移算法实现残余方位空变误差相位估计和补偿。完成以上徙动校正和运动补偿后,对方位全孔径数据进行匹配滤波,得到超高分辨率合成孔径雷达成像结果。采用所提算法对机载X波段和Ku波段实测数据进行成像,得到二维0.05m超高分辨率成像结果,验证了该方法的有效性和实用性。For very high resolution synthetic aperture radar(VHR-SAR), SAR image quality is affected by nonlinear trajectory. To solve this problem, we propose an approach for very high resolution imaging based on motion error estimation. First, inertial navigation system(INS)information is projected on the slant-plane to obtain the motion error, and the coarse range-variant error is compensated by interpolation operation. A modified range migration algorithm(RMA) based on the new Stolt interpolation kernel is proposed and a corresponding motion error estimation method is also proposed to estimate phase errors by the use of sub-aperture error estimation and full aperture error fitting. Then the map-drift(MD) algorithm is used to obtain residual azimuth-variant phase errors estimation. After the above procedures, we obtain the VHR-SAR image with the resolution of 0.05 m, and experimental results of X-band and Ku-band airborne data demonstrate the effectiveness of the proposed method.

关 键 词：超高分辨率 惯导 施托克插值 距离空变 距离徙动算法 

分 类 号：TN957[电子电信—信号与信息处理]