We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
基于 Spark 的合成孔径雷达压缩感知分布式成像.
- Authors
郑灿; 廖可非; 欧阳缮; 谢宁波; 蒋俊正
- Abstract
Compressive sensing (CS) method is widely used in synthetic aperture radar (SAR) imaging. However, it has problems such as long computing time and insufficient scalability of computing power. In order to solve the above problems, a distributed imaging method for SAR compressed sensing imaging based on Apache Spark was proposed. First, the compressed data into row vectors according was divided to each row along the azimuth direction, and then performs distance was distributed parallel reconstruction through the Spark distributed computing platform. The result matrix of range imaging was divided into column vectors according to each column, and then the azimuth distributed parallel reconstruction was performed through the Spark distributed computing platform to complete the SAR compressed sensing imaging. By taking advantage of Sparks memory-based distributed parallel computing, the calculation speed is 1. 9 times that of SAR compressed sensing and 1. 4 times that of MapReduce’s SAR compressed sensing (MR-CS) method. It can be seen that the method in this paper can realize the acceleration of SAR compressed sensing imaging.
- Publication
Science Technology & Engineering, 2022, Vol 22, Issue 10, p4005
- ISSN
1671-1815
- Publication type
Article