Accuracy assessment of height coordinate using unmanned aerial vehicle images based on orthometric height

Abstract

Most of the Unmanned Aerial Vehicle (UAV) images studies are focusing on the horizontal plane with minimal covered on its height values. This thesis will study about the accuracy assessment in height coordinates using UAV images processing based on orthometric height observation data. Apart from that, the accuracy of height coordinates using UAV images processing based on different number of ground control point (GCP) are also assess. In this study, Agisoft PhotoScan v0.9.0 and Global Mapper v15.2.3 are the software used for the processing and analysis. Ground control points and check points were established using Real Time Kinematic Global Positioning System (RTK-GPS) and leveling method. There are two photogrammetric products produced in this study such as orthophoto and digital elevation model (DEM). The analyses of the photogrammetric products were performed based on different number of GCP and different orthometric height observation data. The results of this study showed that the height accuracy of photogrammetric products using UAV images processing based on 10 GCPs produced better accuracy than 5 GCPs. RMSE values of check point height coordinates based on data from leveling method are ± 0.161 meter for 5 GCPs and ± 0.116 meter for 10 GCPs while RMSE values of check point height coordinates based on data from RTK-GPS method are ± 0.167 meter for 5 GCPs and ± 0.124 meter for 10 GCPs. The results also showed that photogrammetric products using UAV images processing based on leveling method provide more accurate height results than RTK-GPS method. The RMSE value of check point height coordinates based on RTK-GPS method is ± 0.148 meter (10 GCPs) and RMSE value of check point height coordinates based on leveling method is ± 0.116 meter (10 GCPs). As conclusion, UAV images can be used to generate DEM that give height coordinate values with sub-meter accuracy.