Towards real-time processing for UAV-mounted GPR-SAR imaging systems

Abstract

Ground Penetrating Radar (GPR) systems mounted on Unmanned Aerial Vehicles (UAVs) are becoming a promising solution to inspect beneath the soil surface. In particular, resorting to Synthetic Aperture Radar (SAR) techniques, these systems are able to retrieve subsurface radar images, in which buried targets can be detected. However, one of the major bottlenecks of this technology is the significant amount of time required for computing the radar images (due to the usage of computationally expensive algorithms to handle non-regular acquisition domains). To face this challenge, this work proposes an efficient interpolation approach, combining a height shift operation with a 2D interpolation, that enables the adoption of a migration algorithm called Phase Shift Migration (PSM). As a result, the processing time is drastically improved, offering close to real-time operation. The proposed methodology has been validated in a realistic scenario for detecting buried explosive targets, comparing it with conventional backprojection techniques in terms of image quality and processing time.