Non-destructive detection and three-dimensional imaging of internal defects in Beijing Ming Great Wall

The Ming Great Wall in Beijing suffers from hidden voids, detachments, and moisture infiltration from long-term weathering and human impacts. Invasive diagnostics risk irreversible damage to this UNESCO site. This study developed a non-destructive 400 MHz GPR framework integrating attribute analysis and 3D imaging, validated through scaled physical models and field surveys at the Panlongshan section. Instantaneous amplitude enabled 3D reconstruction with 19.41% mean volumetric error. Strong empirical correlations (Pearson | r | = 0.79–0.92) link defect moisture content positively to RMS amplitude and generalised S-transform response, and negatively to central-frequency bandwidth and high-frequency (>0.2 MHz) amplitude stability, achieving semi-quantitative moisture discrimination in 13 infill scenarios. While robust in controlled tests, these relationships require site-specific calibration due to material and environmental heterogeneity. The method provides accurate defect localisation and volume estimation for Ming masonry conservation; broader validation across diverse sections is recommended to confirm general applicability worldwide.