Damage Identification and Model Reconstruction of Guangxi Ethnic Architecture Based on Computer Vision Methods

Ethnic architecture serves as a significant cultural symbol and historical treasure of Guangxi, a region densely inhabited by ethnic minorities. This study innovatively integrates computer vision techniques with deep learning models to enhance the precision of damage identification in architectural images acquired through UAV-based oblique photography, which provides more detailed data for model reconstruction, thus realizing a higher accuracy of building model restoration. Furthermore, the methodology enables digital preservation without causing secondary damage to the structures, offering a novel solution for the protection and inheritance of ethnic architecture in Guangxi.