A Novel Optimization Approach for Revolutionizing Architectural Design in Chinese Cultural Heritage

https://doi.org/10.28991/HIJ-2025-06-01-011
Architectural Design Cultural Heritage Building Information Modeling (BIM) Machine Learning Ensemble Ant Colony Fused Convolutional Capsule Neural Network Augmented Reality (AR).

Authors

  • Xueyong Li
    hxiaoqin417@gmail.com
    Fine Arts, Hebei Vocational University of Industry and Technology, Shijiazhuang, 050091,, China
  • Xiaoqing Yang Architectural Engineering, Hebei Vocational University of Industry and Technology, 050091, Shijiazhuang,, China
Vol. 6 No. 1 (2025): March
Research Articles

Downloads

The preservation of China's cultural heritage architecture, which combines contemporary and ancient building techniques, is difficult because of the aesthetic and structural degradation that has overtaken it. This architecture is a testament to the country's technical, artistic, and cultural achievements. A smokescreen with a resolution of 5192 í— 4153 pixels was used to acquire surface photographs and ground shots of the Dazu Rock Carvings, Nanchan Temple, and Foguang Temple using the Microtrans Maryland 4-1000 program. The research aims to improve fault analysis in images of Chinese cultural heritage structures using an Ensemble Ant Colony Fused Convolutional Capsule Neural Network (EAC-CCNN). Then, using a combination of Augmented Reality (AR) and Building Information Modeling (BIM), the designing model for safety management and decision-making will be enhanced. Steps include collecting and annotating data, developing a hybrid EAC-CCNN model to probe the issue with the architectural building, training the model, connecting it with BIM, inspecting the site, and then analyzing the defects using augmented reality (AR) enhanced BIM models. The results show that this integrated approach works to increase the accuracy of defect identification, promote cooperation, and help maintain and preserve cultural heritage assets. The machine learning model's ability to detect and classify defects in buildings that are considered part of China's cultural heritage is evaluated using metrics such as accuracy and F1 score. "With an F1 Score of 95.47% and an accuracy of 93.29%, the architectural design fault identification and safety management model produces respectable results. Phases of training, validation, and testing measure performance in relation to project objectives. Using this approach, machine learning models may be taught to see patterns, fix errors, and make wise predictions under different conditions.

 

Doi: 10.28991/HIJ-2025-06-01-011

Full Text: PDF