MAP RECONSTRUCTION AND LOCALIZATION: A COMPARISON OF LIDAR ODOMETRY ALGORITHMS FOR UAV NAVIGATION

In autonomous UAV navigation, LiDAR-based state estimation, localization,and mapping are necessary for precise obstacle avoidance and 3D mapreconstruction. This paper provides a comparative study of three remarkable LiDARInertial Odometry (LIO) algorithms: FAST-LIO2, LIO-SAM, and LeGO-LOAM. Due tothe generic approach of the algorithms, a comparative analysis of the odometryaccuracy and mapping quality is proposed in this article, which is conducted insimulation environment with a quadcopter UAV equipped with a 3D LiDAR sensor.Overall analysis on odometry accuracy and mapping quality showed great results ontesting the three algorithms. With the dataset tested, most of them showed preciseresults (less than 0.2 m error)