Iteration Phase Gradient Resample Autofocus: A Real-Time and Robust Algorithm for Focusing Maneuvering Ships in SAR Images

Synthetic aperture radar (SAR) is a crucial tool for maritime monitoring. Maneuvering ships often appear defocused in SAR images. Ship refocusing typically employs inverse synthetic aperture radar (ISAR) methods, which are divided into translational compensation and rotational compensation. Currently, rotational compensation lacks robust, efficient, and non-parametric methods akin to phase gradient autofocus (PGA) for translational compensation. This paper proposes an iterative phase gradient resampling autofocus (IPGRA) method, extending the widely used PGA algorithm to compensate for space-variant phase errors caused by non-uniform rotation. The proposed algorithm consists of the following steps: First, extract the defocused ship and perform azimuth decompression and translational motion compensation. Second, divide the echo signal into two parts according to the range cell, use range envelope alignment and PGA to extract the phase error for each part. Third, resample the echo signal based on the phase error. Fourth, introduce a coefficient to describe the degree of rotational defocusing, and repeat the second and third steps until this coefficient converges. Fifth, the Keystone transform is applied to correct the migration through range cells. The proposed algorithm requires only range compression parameters, can achieve undistorted non-sparse imaging, and can quantify imaging quality through the defocusing coefficient. The algorithm inherits the noise resistance and robustness of PGA. Simulation and measured data verify the effectiveness of the proposed algorithm. Finally, this paper introduces an onboard real-time precise focusing processing system incorporating the proposed algorithm.