Identification methods and evaluation metrics for the condition of the Beijing masonry Great Wall

Due to factors such as natural erosion and human interference, the Chinese masonry Great Wall faces challenges in structural stability. In order to solve the problem of low efficiency of manual inspection and lack of evaluation indicators, a multi-source heterogeneous dataset covering more than 500 kilometers of the Great Wall in Beijing was first constructed, and the MEP-Deep disease detection model was proposed. The accuracy of the model on the ISPRS Potsdam dataset and the self-built Great Wall dataset reached 86.37% and 83.51%, respectively, which was 0.6% and 0.77% higher than the original model. Secondly, a quantitative evaluation method for the preservation status of the Great Wall integrating multi-dimensional features was proposed, and a weighted scoring model was constructed using the Analytic Hierarchy Process (AHP). The feasibility of the method was verified experimentally.