Abstract
Unmanned Aerial Vehicles (UAVs) have popularly used as a controllable platform for remote data acquisition. In this study, the DJI spreading wings S800 (DJI S800) was selected as platform to develop an UAV-based system for monitoring the rice crop. The used sensor is Hokuyo UTM30LX laser scanner with 0.37 kg of weight, detection range from 0.1 to 30 m with the wide angle of 270°. The test flight was performed for assessing the accuracy of 3D point clouds at a small and flat experimental site in Kumamoto Prefecture, Japan on Dec. 18th, 2014. Nine ground control points were arranged and their coordinates were identified in the local coordinate system using a total station. By using the GPS data for determining flight direction, the 3D coordinates of point clouds were generated. The location of markers was identified directly from 3D point cloud, and their observed coordinates were manually measured. The 7 parameters transformation was applied with four ground control points. As a result, the 3D point cloud was achieved with the mean absolute error (MAE) of 3.13 m in horizontal and 1.09 m in vertical in cases of non-applying correction method. In case of applying 7 parameters transformation, the absolute error of 3D the point cloud generated from the developed system was improved to 1 m in horizontal and several decimetres in vertical. Although the accuracy of 3D point cloud is not suitable for mapping purpose but is good for the purpose of monitoring the crop in precision agriculture.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call