Pavement drainage pipe condition assessment by GPR image reconstruction using FDTD modeling

Abstract Drainage pipes, as a common water collection system, need to be monitored to ensure proper functionality. One non-destructive way to assess the condition of drainage pipes is to use ground penetrating radar (GPR). The main objective of this study is to develop algorithms to process and interpret the simulated GPR signal from drainage pipes embedded in concrete pavement. This allows locating drainage pipes and predicting their size and condition. The research focuses on reconstructing target images of drainage pipes embedded in concrete pavement. GPR signals were simulated using finite-difference time-domain (FDTD) method. Migration reconstruction and sparse reconstruction methods were used to reconstruct the drainage pipes target. It was found that both methods can predict the condition of the drainage pipe, i.e., whether the pipes are empty or occupied by water or soil. The sparse reconstruction method has better reconstruction accuracy and can be used to locate the drainage pipes and determining pipes diameter. The migration reconstruction method is easier to perform and requires less computational time.