An ISAR Imaging Method Based on Improved CAMP Algorithm

• Published in: IEEE sensors journal Vol. 21; no. 12; pp. 13514 - 13521
• Main Authors: Cheng, Ping, Cheng, Jiawei, Wang, Xinxin, Zhao, Jiaqun
• Format: Journal Article
• Language: English
• Published: New York IEEE 15.06.2021; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Approximation algorithms; Complex approximate message passing (CAMP); compressed sensing (CS); Image reconstruction; Imaging; Inverse synthetic aperture radar; inverse synthetic aperture radar (ISAR); Matching pursuit algorithms; Message passing; off-grid; Optimization; Position errors; Radar imaging; Reconstruction; Scattering; Sensors; Signal to noise ratio; Taylor series
• ISSN: 1530-437X; 1558-1748
• DOI: 10.1109/JSEN.2021.3068281

Compressed sensing (CS) provides a new idea for inverse synthetic aperture radar (ISAR) imaging. Complex approximate message passing (CAMP) algorithm, as a new CS reconstruction algorithm, has characteristics of fast convergence and good reconstruction, which is suitable for ISAR imaging. However, the off-grid problem is inevitable in ISAR imaging, which affects the reconstruction performance seriously and needs to be solved urgently. To solve the off-grid problem, a parametric CS ISAR imaging model is built in this paper, in which the off-grid is considered as an unknown parameter. To solve the joint optimization problem efficiently, first-order Taylor series approximation is used and the optimization problem is transformed into two least squares problems. To verify the effectiveness of the proposed method, it is applied into random sparse signals, complex sinusoids, quasi-real ISAR data and real ISAR data. Compared with the existing algorithms, the proposed method has got higher reconstruction signal-to-noise ratio (RSNR), smaller relative reconstruction error, smaller frequency position error and better imaging result. Therefore the proposed method is an effective CS ISAR imaging method which can solve the off-grid problem well.