Construction of a Machine Learning Based Database for Building Facades: A Case Study of Shantou Xiaogongyuan District

The preservation of historical street styles faces challenges due to the lack of design and restoration guidelines and the limitations of social awareness. Creating a digital record and analysis of buildings with historical value offers an effective approach for preserving cultural heritage in the digital era. Taking historically styled buildings in Shantou's Xiaogongyuan district as an example, a comprehensive database is established in this study by digitally encoding facade elements and training machine learning model. This enables applications such as historical style assessment and prototype analysis of building facades. The study not only provides scientific guidance for the conservation and restoration of historical buildings in cultural districts but also highlights the significant potential of digital technology in cultural heritage preservation.