Prediction of moisture content of brick walls of historical buildings in southern Fujian based on neural network

A high-accuracy neural network model successfully predicts moisture content in brick walls of historical buildings on Gulangsu Island. The model, trained and validated with field data from eight red-brick structures, demonstrated strong generalization (R² values) across training, testing, and verification phases. This approach offers a reliable method for assessing moisture levels in similar heritage sites in southern Fujian, aiding conservation efforts by providing precise predictions without destructive sampling.

Lin, Y., Angkawisittpan, N., Fu, X., & Lyu, G. (2025). Prediction of moisture content of brick walls of historical buildings in southern Fujian based on neural network. Proceedings of the 2nd International Conference on Image Processing, Machine Learning, and Pattern Recognition, 25–31. https://doi.org/10.1145/3759928.3759933