2D Range Flow-based Odometry fusing LiDAR and IMU

Abstract

This paper describes a 2D odometry method fusing LiDAR and IMU data to estimate ground vehicle motion and performing better in long corridor scenarios and moving objecting scenarios. The method builds on range flow constraint for every scanned point and rejects moving objects by fusing IMU data. Besides, our method applies the graphical model to solve the optimization problem that adds IMU constraint to the odometry system. Experiment results demonstrate that our approach is able to perform as well as other state-of-art approaches and better than these approaches in long corridor scenarios and moving objecting scenarios. Besides we test our approach on an embedded system which makes it suitable for those low-cost robotic applications that require planar odometry.