An Autofocus Algorithm of GCBP Image Based on Spectrum Compression for Curved Trajectory Spotlight SAR

The Back-projection (BP) algorithm has significant advantages for spotlight synthetic aperture radar (SAR) imaging with curved trajectory. However, when the motion measurement is not accurate enough, how to autofocus the defocused BP image is still a challenge. In this paper, aiming at the problem of motion error in the application of curved trajectory spotlight SAR, an autofocus algorithm of ground Cartesian BP (GCBP) image based on spectrum compression (SC) for curved trajectory spotlight SAR is proposed. Firstly, through the wavenumber analysis, we reveal the Fourier transform (FT) relationship between the image and the range-compression phase history data under the ground Cartesian coordinate system. Then, the two-step SC of the GCBP image is carried out. And by analyzing the image spectrum, it is found that the spectrum support regions of different targets are aligned and the phase errors of the alignment positions are consistent. Accordingly, efficient error estimation and compensation are carried out. Finally, the simulation experiment of curved trajectory spotlight SAR is given, which proves the effectiveness of the proposed algorithm.