Airborne SAR Motion Compensation and Imaging Based on GPU Architecture

Abstract

A novel and effective parallel airborne SAR imaging method based on GPU architecture is illustrated in this paper. First, characteristic of airborne motion is presented and SAR imaging algorithm is briefly discussed. Second, motion estimation method and robust autofocus algorithm based on GPU are proposed and validated with real data. Third, some final results of SAR imaging are presented and a total time consumption table is listed. The high quality and real-time property of SAR image demonstrate the validation of SAR processing.