IEEE Access (Jan 2022)

The B-Scan Image Simulation Method of a Ground-Penetrating Radar Mounted on a Drone Using a High-Frequency Technique

  • Kittisak Phaebua,
  • Titipong Lertwiriyaprapa,
  • Santana Burintramart,
  • Akkarat Boonpoonga

DOI
https://doi.org/10.1109/ACCESS.2022.3188455
Journal volume & issue
Vol. 10
pp. 71656 – 71668

Abstract

Read online

A B-scan image simulation method of a ground-penetrating radar mounted on a drone using a high-frequency technique is proposed. The proposed simulation method aims to predict drone mounted-GPR receiving signals in a complex three-dimensional (3D) scenario, including the reflected signal from various ground types, backscattered signals from considered underground objects, and backscattered volume clutter signals from underground random object clusters. The modeled scenario, including various 3D drone trajectories, target objects, underground random object clusters, ground types, antenna radiation patterns, and desired transmitting time-domain short pulse signals, are taken into account. A high-frequency technique consisting of the geometrical optic (GO) ray-tracing technique to find the GO ray paths, and the GO ray field approximation technique to find the electromagnetic (EM) fields of every GO ray path, are employed instead of conventional numerical methods. This is useful for EM problems with large computational domains. The efficiency in terms of computational time and computer resources is better than that of numerical techniques. Finally, the effect of drone positioning errors, ground types, underground random object clusters, and the antenna radiation patterns on A-scan and B-scan images of adrone-mounted GPR will be illustrated. It is found that the distortion of the hyperbolic signature of underground target objects in the B-scan image occurred by unwanted back-scattered signals or radar clutter from the ground surface, volume clutter from underground random object clusters, and drone positioning errors. Moreover, a high directivity antenna enhances the intensity of the hyperbolic signature. The proposed simulation method will be useful for predicting the drone mounted-GPR signals in various complex 3D scenarios, having various kinds of transmitting signals, and target object configurations and antenna types.

Keywords