基于分数阶傅里叶变换的运动舰船聚焦算法

Moving Ship Refocusing Using Fractional Fourier Transform Algorithm

为解决合成孔径雷达(SAR)图像中运动舰船目标产生的散焦现象,结合对比度最大算法和分数阶傅里叶变换(FRFT)算法,提出了一种改进的对比度分数阶傅里叶变换(CFRFT)自聚焦算法。该算法利用分数阶傅里叶变换对已成像SAR图像进行时频域分析,根据旋转角分别利用参数模型和非参数模型对二阶相位误差和高阶相位误差进行补偿,和传统的相位梯度(PGA)法相比,图像分辨率和旁瓣比提升显著,可以更有效地补偿SAR中舰船运动产生的相位误差。对不同舰船和尾迹SAR图像实验表明,算法对二阶以上的相位误差具有较好的补偿效果,误差估计准确性高,适用范围广,解决了SAR运动舰船的散焦问题,提高了海洋舰船监测的准确性。

In order to refocus the image of moving ship in synthetic aperture radar （SAR）,a detection ap-proach called contrast fractional Fourier transform（ CFRFT） is presented. This algorithm is used to process SAR image by using contrast maximum and fractional Fourier transform （ FRFT） method. Based on FRFT method,the signature can be represented in a rotated jointed time-frequency plane and the optimal rotated degree can be found. Taking advantage of the optimal degree,the quadratic phase error is estimated by a parametric model processing and the residual phase error can be obtained by a nonparametric model pro-cessing. The defocusing SAR image can then be compensated easily. Compared with traditional phase gra-dient autofocus（ PGA） algorithm,the CFRFT can compensate the defocusing caused by moving ship target more effectively. The resolution and side lobe rate are also improved. The simulation and the processing results of moving ship show that this algorithm can get well focused SAR images with robustness,effectivity and high precision.