Evaluasi Performa DEM Sensor Visual (RGB) terhadap DEM Sensor LiDAR melalui Perbandingan Elevasi dan Volume Sedimen di Bendung Kamijoro

Sediment mapping is an important aspect of dam management because sediment accumulation can reduce storage capacity and interfere with intake performance. This study aims to compare the performance of Digital Elevation Models (DEM) derived from visual sensors (RGB) with LiDAR DEMs through elevation difference analysis and sediment volume estimation at the Kamijoro Dam. Data were obtained through Unmanned Aerial Vehicle (UAV) surveys at a flight altitude of 120 m, then processed to generate DEMs from each sensor. The analysis was conducted by calculating the elevation difference at the comparison points and comparing the estimated sediment volumes produced. The results showed that the visual DEM was able to represent the general surface elevation pattern, but showed greater elevation variation compared to the LiDAR DEM. The minimum elevation difference recorded was −4.3046 m, the maximum was 9.4975 m, with an average value of 1.0323 m. In the sediment volume estimation, the LiDAR DEM produced a volume of 85,612.456 m³, while the visual DEM produced 128,596.23 m³. This difference indicates a tendency for overestimation in the visual DEM, which is thought to be caused by limitations in separating the ground surface from vegetation and non-surface objects. Relative to the LiDAR DEM, these results show that the visual DEM has higher elevation variability in complex surface conditions.