SAR imaging method, device and system

Abstract

The invention discloses an SAR imaging method including: dividing received first full-aperture SAR echo data into more than two sections of sub-aperture SAR echo data in a azimuth direction and carrying out distance-direction processing on the more than two sections of sub-aperture SAR echo data; combining the more than two sections of sub-aperture SAR echo data which undergo the distance-direction processing into second full-aperture SAR echo data; dividing the second full-aperture SAR echo data into more than two SAR echo-data blocks in the distance direction and processing the more than two SAR echo data blocks in the azimuth direction; and combining the more than two SAR echo-data blocks which undergo the azimuth direction processing into an SAR image. The invention also discloses an SAR imaging device and system. Through adoption of the scheme, high-speed and high-quality imaging processing on the SAR echo data in a large scale is realized with an excellent imaging effect and GPU speed-up ratio.

Description

technical field [0001] The present invention relates to synthetic aperture radar (SAR, Synthetic Aperture Radar) imaging technology, in particular to a SAR imaging method, device and system. Background technique [0002] SAR is an active earth observation technology. Compared with traditional optical sensors, SAR can realize all-time and all-weather real-time observation of the earth, and has certain surface penetration capabilities. The development of SAR technology has brought about a sharp expansion of the amount of data and calculations, and with the increasing demand for real-time imaging, the real-time imaging processing of large-scale SAR echo data is becoming more and more important. [0003] In recent years, with the continuous improvement of the parallel processing capability of graphics processing units (GPUs, Graphic Processing Units), GPUs are being increasingly used in scientific computing fields, such as image segmentation, SAR echo simulation, etc. In the fi...

AI Technical Summary

Problems solved by technology

However, as the SAR echo data increases, the GPU cannot complete the SAR echo processing of the entire scene at one time due to the limitation of the GPU memory capacity.

If the traditional imaging algorithm is used, it is necessary to continuously transmit large-scale data between the central processing unit (CPU, Central Processing Unit) and the GPU, while the new generation bus interface (PCI-E, Peripheral Component Interconnect-Express) The transmission bandwidth is very small compared to the memory read bandwidth of the GPU, so this will seriously restrict the processing speed of the GPU, so for large-scale SAR echo processing, it is far from enough to implement the original imaging algorithm in parallel on the GPU of

Images