Accuracy Assessment for Geometric Features Extraction from Multirotor UAV's Images

Abstract

The use of the new technology Unmanned Aerial Vehicle (UAVs) is overgrowing across many application fields, including engineering surveys, terrain mapping, or generating 3D models. Accuracy assessment for object extraction from UAV images is critical for a large number of applications and plays an essential role in ensuring its quality. This study aims to assess the accuracy of geometric features that can be acquired with low-cost multirotor UAV. In this study, the UAV used to capture the ground at different altitudes, as the accuracy is influenced by flight height---the data processed by using photogrammetric software, which is based on structure from motion. The accuracy assessment is conducted by comparing the estimated results with those obtained via field measurements. It was found that high altitude gives a better result as compared to low altitude. In conclusion, this study demonstrates the feasibility of using high accuracy UAV's image to provide data for geometric design with reliable accuracy.