Title Reliability Assessment of Photogrammetry-Based 3D Models for Digital Management of Cultural Heritage
Authors 고명환(Ko, Myong-Hwan) ; 강석진(Kang, Seok-Jin)
DOI https://doi.org/10.5659/JAIK.2026.42.7.51
Page pp.51-62
ISSN 2733-6247
Keywords Cultural Heritage; UAV Photogrammetry; Reliability of 3D Numerical Models; Cloud-to-Cloud(C2C) Distance Analysis; Vertical Projection-based Height Difference; Digital Heritage Management
Abstract This study evaluates the reliability of UAV photogrammetry?based 3D models for the management of built cultural heritage exposed to increasing hydro-meteorological hazards. Based on a reliability assessment framework integrating accuracy, precision, and consistency (repeatability and reproducibility), experiments were conducted at Daanseong Hyanggyo (Myungryundang Hall), Korea. Five photogrammetric models were generated using Pix4D Matic and analyzed in CloudCompare. Accuracy was assessed through vertical projection?based height difference (|ΔZ|) analysis using GNSS and total station reference data, while precision was evaluated using check-point residual RMSEs from repeated model reconstructions. Consistency was examined through Cloud-to-Cloud (C2C) distance analysis under varying acquisition conditions. Results show stable vertical accuracy on ground surfaces (0.07?0.11m), while larger local errors occurred at roof edges and ridge elements. Precision and consistency remained stable overall, with 86?95% of points within 0.20m, indicating high repeatability and reproducibility. Comparative analysis with LiDAR data also demonstrated high geometric agreement despite differences in acquisition methods. These findings suggest that UAV photogrammetry provides reliable 3D data for cultural heritage monitoring and change detection, although careful interpretation is required for complex or occluded elements.