3D digital modelling and identification of pavement typical internal defects based on GPR measured data

A three-dimensional ground-penetrating radar (GPR) captures non-destructively internal pavement distress characteristics. However, interpreting radar images and data analysis pose challenges. To improve the accuracy of distress identification, a three-dimensional digital model of internal pavement distress was established. Firstly, initial electromagnetic signal data were pre-processed to effectively eliminate spurious signals and enhance distress characteristic signals. The distress was located, and GPR images of typical distress were extracted and summarised. Next, the 3D dataset was constructed based on the pre-processed electromagnetic echo signals. A 3D digital model of internal pavement distress was generated using the inverse distance weight and ray-casting methods with trilinear interpolation. Finally, relying on the physical project, cores were extracted to validate the distress model. The method effectively reflects the internal pavement distress, and enables realise the interactive images between the pavement entity and the digital model, which can essentially contribute to the digital twin of pavement systems.