Research on the use of UAV-inclined photography and 3D laser scanning in the digital conservation of historic structures: A case study of Shaoxing Lvfu

• Ground-Air Cooperative Mapping System: This study integrates 3D laser scanning (FARO Focus Premium) and UAV tilt photography (DJI Mavic 3E) to achieve high-precision (error <3 mm) and non-contact digital conservation of ancient buildings, addressing limitations of traditional methods. • High-Precision 3D Digital Twin: By combining point cloud alignment (FARO Scene), data fusion (Auto Recap), and BIM conversion, a detailed model with a 58% overlap rate is created, capturing complex geometric and non-geometric features (e.g., materials, colors). • Comprehensive Data Fusion: Merges UAV-derived façade/roof data with laser-scanned interior details, compensating for individual technique gaps (e.g. omitted columns or roof sections), enhancing model completeness. • Practical application: 2D orthogonal projections (scale accuracy 1:100) exported from Revit are utilized for restoration guidance, supporting cultural heritage conservation and public engagement. • Efficiency and Safety: This method reduces fieldwork time (2 million points/second scanning) and eliminates physical contact risks, providing a scalable solution for heritage documentation. Responding to traditional surveying and mapping issues, such as their low accuracy and damage to the contact when protecting ancient buildings, this study uses Shaoxing Lvfu, a nationally recognized key cultural relic site, as the research object and creates a three-dimensional digital model based on ground-air collaborative surveying and mapping of 3D laser scanning and unmanned aerial vehicle oblique photography. The FARO Focus Premium 3D laser scanner (ranging error ±1 mm, sampling rate 2 million points/second) is used to complete the high-density point cloud acquisition of the internal and external structures of the building, and combined with the DJI Mavic 3E UAV tilt camera system (vertical overlap rate of 90%, multi-view image coverage), the multi-dimensional acquisition of the building façade and the roof is realized. Through the FARO Scene point cloud alignment, the error of the 3D laser point cloud model is controlled within 3 mm, and the overlap rate reaches 58%. The 2D orthographic projection map (scale accuracy of 1:100) exported through the Revit model provides a quantifiable engineering basis for cultural heritage restoration. The verticality inspection for the vertical structure of Lvfu shows that its overall verticality deviation is significant, and there is an urgent need to implement long-term structural health monitoring. The technical chain of “data acquisition-integration processing-application transformation” established in this study provides an innovative paradigm of both efficiency and precision for the digital protection of cultural heritage, and its applicability to engineering has been verified by the actual case of Lvfu.