A Novel Compressive Sensing Algorithm for SAR Imaging

• Published in: IEEE journal of selected topics in applied earth observations and remote sensing Vol. 7; no. 2; pp. 708 - 720
• Main Authors: Dong, Xiao, Zhang, Yunhua
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 01.02.2014; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Azimuth; Chirp-scaling algorithm (CSA); compressive sensing (CS); Image reconstruction; Imaging; orthonormal basis; Radar; Radar polarimetry; Sparse matrices; Synthetic aperture radar; synthetic aperture radar (SAR); two-dimensional double CS Algorithm (2-D-DCSA); Vectors
• ISSN: 1939-1404; 2151-1535
• DOI: 10.1109/JSTARS.2013.2291578

A novel compressive sensing (CS) algorithm for synthetic aperture radar (SAR) imaging is proposed which is called as the two-dimensional double CS algorithm (2-D-DCSA). We first derive the imaging operator for SAR, which is named as the chirp-scaling operator (CSO), from the chirp-scaling algorithm (CSA), then we show its inverse is a linear map, which transforms the SAR image to the received baseband radar signal. We show that the SAR image can be reconstructed simultaneously in the range and azimuth directions from a small number of the raw data. The proposed algorithm can handle large-scale data because both the CSO and its inverse allow fast matrix-vector multiplications. Both the simulated and real data are processed to test the algorithm and the results show that the 2-D-DCSA can be applied to reconstructing the SAR images effectively with much less data than regularly required.