Ground penetrating radar attributes analysis for detecting underground artificial structures in urban areas, Vietnam

Abstract

Maps of underground construction works for maintenance, building and planning tasks are essential in urban areas. Ground penetrating radar (GPR) has shown effectiveness by providing high-resolution imaging of subsurface structure. Reflection/ diffraction from boundaries, singular objects, or edges of anomalies could lead to meaningful events like strong signals or shapes of hyperbolae in the common offset unmigrated GPR sections. In urban areas, although the shielded antennas help the GPR machine to remove most of man-made noises transmitted from the air, the GPR data are still heavily contaminated by types of noises. Concerned objects (i.e., underground pipes, tunnels, and subsurface structures) can be effectively revealed by applying the GPR attributes that are calculated from the processed abundant GPR data information as enhancement of ratio between signal over noise. In our paper, for the GPR data surveyed in Ho Chi Minh City, Ba Ria – Vung Tau Province, and Dong Nai Province, Vietnam, we have utilised a set of GPR attributes (i.e., amplitude before/after migration, cosine-of-instantaneous phase, and energy) and volume rendering (i) to define 2D/3D anomalies relating to underground manmade system, and (ii) to uncover hidden underground objects with illumination of upper concrete construction steel net of buildings.