Dual-Path Framework Analysis of Crack Detection Algorithm and Scenario Simulation on Fujian Tulou Surface

A dual-path framework combining advanced crack detection and physical simulation has been developed to address the vulnerability of Fujian Tulou to environmental and human-induced stresses. The study introduces an improved YOLOv12 model for lightweight, accurate crack recognition across complex wall textures, while a two-layer RFPA3D model simulates crack evolution under varying conditions, aligning with field observations. Quantitative validation demonstrates superior consistency in morphology, geometry, and topology compared to baseline models. This framework offers a practical solution for standardized crack documentation and mechanistic interpretation, advancing preventive conservation efforts and sustainable management of Fujian Tulou.

Hu, Y., Chen, S., Zhao, Z., & Cheng, S. (2025). Dual-Path Framework Analysis of Crack Detection Algorithm and Scenario Simulation on Fujian Tulou Surface. Coatings, 15(10), 1156. https://doi.org/10.3390/coatings15101156