Oil palm tree height detection from UAV images.

Traditional methods like counting and measuring the height of the trees manually are time-consuming and also costly, which can be a challenge in the oil palm industry. Therefore, having an automatic system to count trees and measure their height in an oil palm plantation is useful in production planning and decision-making. Using Unmanned Aerial Vehicles (UAVs) to count trees and measure their height has become more popular due to its advantages in ecosystem and forestry industries. In this study, some image processing techniques are used to process the dataset captured by UAV in order to extract the height and number of oil palm trees from the dataset. These datasets is captured using a MicaSense RedEdge Multispectral Sensor that is attached to a UAV that flying 80m above sea level. The processed datasets include the Canopy Height Model (CHM), Digital Surface Model (DSM), and Digital Terrain Model (DTM). This study shows relatively good accuracy between the predicted and prepared CHM where R-squared (R2)=0.74, Mean Absolute Error(MAE)=0.28, and Root Mean Square Error(RMSE)=0.38. [ABSTRACT FROM AUTHOR]