Migration imaging processing of subgrade cavity GPR detection based on TUFK method

Timely and accurate detection and identification of cavities under urban roads is the key to road safety. Due to the inability to obtain accurate migration velocity and the difficulty of achieving complete convergence of diffraction signals of the cavity disease, traditional migration methods struggle to accurately identify and locate the subgrade cavity. This paper proposes a GPR image migration processing (TUFK method) based on 2D undecimated wavelet transform and the F–K method in accordance with the high-precision imaging of the subgrade cavity. The finite-difference forward models of subgrade cavity without and with noise are established, and the model test of cavity detection by GPR is carried out in the laboratory. Through the fine extraction and migration processing of the weak diffraction signals from the cavity, the optimal velocity required for migration is analyzed, and the TUFK method is applied to the migration process of GPR data acquired for the purpose of cavity detection. Furthermore, the proposed method is applied to the processing of GPR data acquired in the field with a cavity below the roadbed. The results show that the TUFK method can accurately extract the diffraction signals from the cavity and achieve the fine convergence of cavity diffraction signals whether in noiseless or noisy environments. Compared with the traditional Kirchhoff and F–K migration methods, this method can effectively obtain accurate migration velocity and the migration results can reflect the actual position and shape of the cavity. This study can provide a new idea and effective method for the imaging of subgrade cavity.