Non-destructive testing research on the surface damage faced by the Shanhaiguan Great Wall based on machine learning

A machine learning approach significantly enhances the detection of surface damage on the Shanhaiguan Great Wall, a UNESCO-listed Ming Dynasty structure. By applying the YOLOv4 model to analyze gray brick surfaces, researchers achieved a 0.17% increase in accuracy over traditional methods, identifying four types of damage—chalking, plants, ubiquinol, and cracking—without manual intervention or destructive testing. This innovation accelerates monitoring, reduces labor costs, and enables timely repairs while preserving the historical integrity of ancient brick walls.

Li, Q., Zheng, L., Chen, Y., Yan, L., Li, Y., & Zhao, J. (2023). Non-destructive testing research on the surface damage faced by the Shanhaiguan Great Wall based on machine learning. Frontiers in Earth Science, 11. https://doi.org/10.3389/feart.2023.1225585