A Prior 2-D Autofocus Algorithm With Ground Cartesian BP Imaging for Curved Trajectory SAR

• Published in: IEEE journal of selected topics in applied earth observations and remote sensing Vol. 17; pp. 2422 - 2436
• Main Authors: Lou, Yishan, Lin, Hao, Li, Ning, Xing, Mengdao, Wang, Junrui, Wu, Zhixin
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 2024; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Cartesian coordinates; Cell migration; Curved trajectory; Error correction; ground cartesian back-projection (GCBP); Image resolution; Imaging; Imaging techniques; index grid division algorithm; Interpolation; Phase error; prior structure; Radar imaging; Signal processing algorithms; Signal resolution; spectrum alignment; Synthetic aperture radar; synthetic aperture radar (SAR); Trajectories; Trajectory; two-dimensional (2-D) phase error
• ISSN: 1939-1404; 2151-1535
• DOI: 10.1109/JSTARS.2023.3346942

Due to the complexity of the trajectory, the curved trajectory SAR signal is seriously two-dimensional (2-D) space-variant. The ground Cartesian back-projection (GCBP) algorithm is a fast time-domain algorithm without interpolation, which can meet the imaging accuracy of curved trajectory SAR. However, when the nonsystematic range cell migration of GCBP image caused by motion error exists, the existing autofocus algorithms find it difficult to correctly estimate the phase error. Therefore, an extended 2-D autofocus algorithm is proposed to realize curved trajectory SAR imaging by combining the GCBP algorithm, in this article. The expression of the spectrum support region of the GCBP image is given, and the spectrum characteristics are analyzed. Then, the spectrum alignment is performed based on these characteristics to eliminate spectrum aliasing. Finally, the analytical property of phase error in the 2-D frequency domain of the GCBP image is analyzed in detail. Based on the property, the prior 2-D phase error structure for the GCBP image is established. According to the structure, the 2-D phase error is computed and compensated to obtain the refocused image. In addition, the index grid division method embedded in the proposed algorithm can effectively compensate the space-variant phase error. The simulation and real data results prove the effectiveness of the proposed algorithm.