Interpretation of ground penetrating radar image using digital wavelet transform

Abstract

The ultra wide bandwidth nature of ground penetrating radar antenna has made a raw data acquired with the tool prone to unwanted noise and hence low signal to noise ratio. Quantitative interpretation of the data is desirable for radar image quality enhancement. This study used multiresolution analysis to process radar image at different levels of decomposition. Daubechie wavelets family was used to decompose the image into 4-different levels of details. Level 3 diagonal details and level 4 horizontal details provide a noise-free visualization of the subsurface discontinuities. This led to the detection of buried utility and unique identification of its spatial location. The depth of the buried utility was estimated based on the image scale. The work demonstrates the effectiveness of Daubechie wavelet transform analysis as yet another technique of utility survey data acquired with GPR.