해안사구 변화 모니터링을 위한 시계열 UAV DEM 표고보정

다국어 초록 (Multilingual Abstract)

Due to climate change and coastal development, significant transformations in coastal dunes have increased the demand for continuous monitoring. The usage of UAV (Unmanned Aerial Vehicles) for this purpose has escalated. However, early UAV monitoring methods encountered limitations in the continuous installation and field measurement of GCP (Ground Control Points), necessitating the use of a reference DEM (Digital Elevation Model) for elevation correction. Notably, as the study area includes coastal dunes on islands, which differ from the mainland, long-term monitoring involves substantial costs and manpower for GCP surveying. This research introduces an elevation correction method for monitoring changes in coastal dunes, building upon previous studies. UNET was used to distinguish classes in invariant areas and to extract characteristic values from these areas. These values were normalized using the Z-score to eliminate outliers, and a linear regression formula was applied to adjust the DEM values. Over eight years, 13 sets of orthoimages and DEMs were created, both with and without GCPs. The proposed technique was evaluated by comparing DEMs with GCPs to those corrected using the proposed method without GCPs. Additionally, transects were selected to assess the effectiveness of the proposed monitoring method. This study confirms the feasibility of using the corrected DEM without ground control point data for coastal dune monitoring.