基于最优对比度准则的SAR图像相位梯度自聚焦算法

相位梯度自聚焦（PGA）算法是SAR图像自聚焦领域的经典算法。在分析经典PGA算法性能的基础上，提出了基于最优对比度准则的SAR图像相位梯度自聚焦算法。该算法改进了PGA算法的加窗方法和迭代终止条件，提高了PGA算法的鲁棒性和实时应用性能。理论分析和实验结果证明了该算法的有效性。

Phase Gradient Autofocus Algorithm for SAR Imagery Based on Contrast Criteria

Phase gradient autofocus(PGA) algorithm is a well known method for the estimation and the compensation of phase errors in synthetic aperture radar system,which relies on the strong and isolated scattering point in each range line.But,employing the range lines possessing low contrast can deteriorate at a certain extent and even invalidate the phase gradient autofocus algorithm.In addition,the dominating defaults of this algorithm are the sensitivity to the windowing width and high computational complexity.By analyzing the relation between the number of range line and the energy distribution in azimuth,we mend the choice method of the window width and the terminated condition of the iterative algorithm,basing on the optical contrast criteria.The simulated data are also exploited to investigate the effectiveness of the proposed method.Theoretical analysis and the experimental results all show that the improved phase gradient autofocus algorithm not only can remain the convergence of the iterative algorithm,but also can reduce the computational burden dramatically.Further more,the proposed method improves largely the robustness of the algorithm.All these make it possible to applying the modified phase gradient autofocus algorithm to the high resolution synthetic aperture radar system at real time.