一种用于条带模式SAR成像的自聚焦算法

A Novel Approach to Strip-map SAR Autofocus

针对现有条带模式合成孔径雷达(SAR)成像的自聚焦算法中,相位的拼接会引起误差的严重积累,本文结合相位梯度自聚焦(PGA)算法和子孔径偏移(MD)算法,提出了一种新的用于条带模式SAR成像的自聚焦算法(PGA-MD)。该算法先利用PGA算法有效估计各子孔径相位误差函数,再利用相邻子图间方位向偏移量和线性相位之间的关系,通过MD算法精确估计相邻子图间方位向偏移量,然后计算线性相位差,并基于—阶导数实现子孔径相位误差函数拼接。理论分析以及实测数据处理结果对比均表明PGA-MD算法可以有效提高对条带SAR图像的自聚焦效果。

For the dramatic accumulation of estimation errors during the procedure of the subaperture phase error combining in the persent techniques of autofocus for the conventional strip-mapping mode synthetic aperture radar（SAR）imagery,this article proposes a novel approach to strip-map SAR autofocus,which combines the techniques of phase gradient autofocus（PGA）and map drift（MD）.The proposed method also requires that the complete procedure of PGA be accomplished for each subaperture fist.In order to decreasing the accumulation of estimation errors during the procedure of the subaperture phase error combining,the basic rationale of the presented PGA-MD is to adjust and combine the first difference of the subaperture phase error via MD,which can achieve an accurate estimate of the azimuth shift between the neighboring sub-images.Comparison on the accuracy of this approach against the traditional algorithm is presented.Experimental results indicate that the presented PGA-MD methodology can improve the quality of strip-map SAR image effectively.