Quantitative estimation of asbestos-containing slate roof areas using UAV imagery and GIS-based correction

초록

This study proposes an integrated methodology that combines unmanned aerial vehicle (UAV)based high-resolution imagery, 3D modelling, and geographic information system (GIS) analysis to accurately quantify asbestos cement slate roof areas. Conventional field surveys are limited by accessibility, subjectivity, and inefficiency, especially for aging or structurally vulnerable buildings. To overcome these challenges, aerial images were acquired using a UAV, followed by the generation of orthomosaics and digital elevation models (DEMs) using Pix4Dmapper. Roof slopes were extracted in the ArcGIS environment to calculate geometrically corrected roof areas, and a 16 % overlap ratio was applied to reflect the material's overlapping installation characteristics. The methodology was applied to 45 buildings, and results demonstrated a high level of agreement with expert visual inspections, with an average error of approximately 8.5 m2, a root mean square error (RMSE) of 18.61 m2, and 68 % of cases falling within a +/- 10 m2 error range. These findings indicate that the proposed UAV-based approach enables repeatable, accurate, and scalable quantity estimations, with potential for practical applications in asbestos abatement planning, environmental risk assessment, and urban regeneration. This study also highlights opportunities for future enhancement through AI-based roof recognition, automatic boundary detection, and multi-sensor integration.

키워드