22 April 2015 Highly efficient synthetic aperture radar processing system for airborne sensors using CPU+GPU architecture
Author Affiliations +
Abstract
A comprehensive and effective parallel airborne synthetic aperture radar (SAR) processing system using the architecture of central processing units (CPU) and graphical processing units (GPU) is proposed first in this paper. By using the techniques of Compute Unified Device Architecture, the SAR processing system is much more efficient and robust, thereby enabling it to work with high efficiency. This paper focuses on the optimizations of motion compensation, subaperture chirp scaling algorithm, phase gradient autofocus (PGA), and even visualization with Open Source Computer Vision Library. Apart from this, it is also the first time the PGA has been run on the architecture of CPU and GPU. Experimental results show a speedup of ∼37 times compared with a nonoptimized CPU-based approach.
© 2015 Society of Photo-Optical Instrumentation Engineers (SPIE)
Zheng Wu, Yabo Liu, Lei Zhang, Ning Li, Kangning Du, Timo Balz, "Highly efficient synthetic aperture radar processing system for airborne sensors using CPU+GPU architecture," Journal of Applied Remote Sensing 9(1), 097293 (22 April 2015). https://doi.org/10.1117/1.JRS.9.097293 . Submission:
JOURNAL ARTICLE
12 PAGES


SHARE
RELATED CONTENT

SAR imaging and detection of moving targets
Proceedings of SPIE (September 15 1998)
Image exploitation for MISAR
Proceedings of SPIE (May 07 2007)
SAR image formation processing using planar subarrays
Proceedings of SPIE (June 09 1994)
DARPA interest in diffractive sensors
Proceedings of SPIE (September 26 2007)

Back to Top