jaue2023061: A Study on Establishing a Quantification Process for Asbestos Slate Using UAV Image-Based Methods

Abstract

In this study, we select a target site where asbestos slate quantification has been completed and explain how to identify and verify the quantity of asbestos slate through aerial photography using a UAV. Autonomous enroute flights were performed based on pre-planned altitudes, routes, and repetitions. Using the Pix4D mapper program, we constructed an orthoimage and digital elevation model (DEM) through image adjustment, match point extraction, point cloud creation, and mesh creation processes. A total of 72,695 matching points were extracted during the image adjustment stage, ensuring a quality level of GSD of 2.62 cm per pixel. Based on the generated DEM, the slope of the asbestos slate roof surface was measured using ArcMap, and a polygon was created based on the roof slope to measure the slope of each roof surface. As a result of comparing the building ledger and integrated actual measurement data, the average error was about 8.5m2, the maximum error was 89.03m2, and the minimum error was 0.06m2. By applying the methodology presented in this study, an asbestos slate quantification process using UAV was established.
If the process developed in this study is used to investigate asbestos slate, it will be able to provide basic data for building an asbestos slate database. This database includes annotated images acquired via UAVs, geo-referenced orthoimages, and asbestos slate quantities and conditions.