In-Situ Calibration and Trajectory Enhancement of UAV LiDAR Systems for Mapping Mechanized Agricultural Fields

Uncrewed aerial vehicles (UAVs) carrying sensors, such as light detection and ranging (LiDAR) and multiband cameras georeferenced by an onboard global navigation satellite system/inertial navigation system (GNSS/INS), have become a popular means to quickly acquire near-proximal agricultural remote sensing data. These platforms have bridged the gap between high-altitude airborne and ground-based measurements. UAV data acquisitions also allow for surveying remote sites that are logistically difficult to access from ground. With that said, deriving well-georeferenced mapping products from these mobile mapping systems is contingent on accurate determination of platform trajectory along with intersensor positional and rotational relationships, that is, the mounting parameters of various sensors with respect to the GNSS/INS unit. Conventional techniques for estimating LiDAR mounting parameters (also referred to as LiDAR system calibration) require carefully planned trajectory and target configuration. Such techniques are time-consuming, and in certain cases, not feasible to accomplish. In this article, an in-situ system calibration and trajectory enhancement strategy for UAV LiDAR is proposed. The strategy uses planting geometry in mechanized agricultural fields through an automated procedure for feature extraction/matching and using them to enhance the quality of LiDAR-derived point clouds. The proposed approach is qualitatively and quantitatively evaluated using calibration datasets as well as separately acquired validation datasets to demonstrate the performance of the developed procedure. Quantitatively, the accuracy of the resulting UAV point clouds after system calibration and an accompanying trajectory enhancement improved from as much as 43 to 4 cm.