Application of continuous wavelet transform for ground penetrating radar data interpretation in detection of underground anomaly

Abstract

Ground penetrating radar (GPR), known as a powerful geophysical technique, can detect underground artificial objects in urban environments. The shape of reflected or diffracted signals in GPR data slices can exhibit their locations in 2D/3D space. Continuous wavelet transform is effectively applied in locating and highlighting anomalies in geophysical methods, such as data earthquake analysis, seismic exploration, and gravity studies, by analyzing energy spectra in the time - frequency domain. This study presents a novel approach to GPR data interpretation that integrates continuous wavelet transform with traditional processing techniques. This workflow was applied to both modelled and field data sets. In the modelled case, a circular pipe with an electrical permittivity different from its surroundings was successfully recognized in the time - frequency domain after applying continuous wavelet transform. For the real case, two 2D GPR profiles were conducted to investigate an elongated underground drainage pipe on Nguyen Van Cu Street in District 5, Ho Chi Minh City, Vietnam. 3D representation of the continuous wavelet transform output for the real data demonstrated the presence of the anomaly, characterized by strong energy amplitudes linked with reflection and diffraction events.