The inversion GPR signal by compressing sensing and its application

GPR signal is essentially a delayed, scaled and distorted version of the transmitted pulse, determined by sparse parameters such as time delays and amplitudes. The time delay indicates the detected target's depth, and the amplitude is the impedance contrast at interface or reflection coefficient. A system for sampling GPR signal at an ultra-low sampling rate based on the Finite Rate of Innovation (FRI) and the estimation of time delays and amplitudes to detect GPR signal is presented in the paper. We build the algorithm and apply it to real GPR data from Northwest of China. The results demonstrate that the inverted impedance and time delay can be used to depict the sandstone and clay layer with high accuracy.