High-resolution 3D image reconstruction based on multistatic UWB phase processing

Abstract

This paper presents the first results on multistatic ultra wideband (UWB) ground penetrating radar (GPR) interferometry. It has been known that phase data may contain subtle information of the physical and geometrical properties. Most GPR antennas are used in near-field and/or Fresnel regions; the energy scattered from randomly rough surfaces appears diffusely and merely depends on the wavelength of the incident radiation and the incident angles. The random nature makes it a difficult task to determine the profile of the surfaces. Furthermore it is too time-consuming to solve this kind of three-dimensional inhomogeneous problems by using pure electromagnetic approaches, especially when real time applications are required. Therefore we initiated the research on near-field UWB GPR interferometry to profile randomly inhomogeneous rough surfaces as well as their shallow layers in three-dimensional space. In this paper we discuss the first results based on practically measured multistatic UWB phase data scattered from curved objects. An initial algorithm that was developed for image reconstruction in the surface height domain (SHD) has been applied to the measured data. The image patterns show good agreement with the geometries of the practical curved objects. The results demonstrate the applicability of the phase information to characterization of the curved objects.