[포스터] CREATION OF RIVER TERRAIN DATA USING REGION GROWNING METHOD FROM UAV PHOTOGRAPHY |
---|
학술지명 InternationalSymposium on RemoteSensing2018(ISRS)
저자 유완식
발표일 2018-05-10
|
Terrain data for the areas around rivers are essential not only for developing various comprehensive plans for rivers, including national rivers, local rivers, and small rivers, but also for setting restrictions on indiscriminate development within a target river’s environment and for performing disaster management tasks, such as flood control. Recently, many studies to create three-dimensional terrain data using unmanned aerial vehicles (UAVs) have actively been conducted in the spatial information field. This study aimed to create a DEM using a region growing method after classifying ground points by applying a slope technique based on three-dimensional DSM data from a UAV. First, we selected a part of Gap River situated in Seo-Gu, Daejeon, as a target area for acquiring images using a UAV, and created orthomosaic and DSM data by matching the images. Because the created DSM data included vegetation and trees in the river, we applied a slope technique based on point cloud data, acquired during UAV-based image capturing, to create a DEM based on the ground of the river. The study also aimed to evaluate the elevation accuracy of the DSM and DEM data by selecting 100 test points and using the measurement values calculated based on VRS and total station surveys, in order to demonstrate the reliability of the DEM data created with a slope technique using point cloud data from a UAV. As a result, the standard errors of DSM and DEM were ± 3.84m and ± 0.12m, respectively. Therefore, it was confirmed that the accuracy of DEM can be secured by applying the UAV point cloud based on region growth algorithm, and it can be effectively used for the river management tasks. |