改进的相位梯度自聚焦加窗方法

相位校正是ISAR（逆合成孔径雷达）成像中的关键步骤，校正的精度直接关系到成像的质量。相位梯度自聚焦算法是一种广泛使用的相位校正算法，可以有效地补偿目标复杂运动所造成的相位误差。本文在研究传统PGA（相位梯度自聚焦）算法的基本步骤所起作用的基础上，对加窗方法进行改进，提出以归一化平均能量来估计误差支撑区域的方法。该方法综合考虑了非目标区域与目标区域的归一化平均能量，并将两者的均值作为提取误差支撑区域的阈值。对实测数据进行成像的处理结果表明，该方法可以提高自聚焦算法的收敛速度，使其更具有鲁棒性和实时性。

An Improved Method to Select Window Width for the Phase-gradient Autofocus Algorithm

As one of the key steps in inverse synthetic aperture radar （ ISAR）, phase compensation can affect the quality of the image. Phase-gradient autofocus （PGA） algorithm is a very popular way that can effectively compensate the phase error caused by complicate movement of the target. Based on the analysis of the basic steps of the phase-gradient autofocus algorithm, an improved way to select window width is proposed, which uses the unitary average power to estimate the error-support-area. The new method takes the unitary average power of the target-area and background into account, using the average value as the threshold to estimate the error-support-area. The imaging tests with real ISAR data demonstrate that this method can improve the convergent speed and enhance the robustness and efficiency of the algorithm.