SW-LIO: A Sliding Window Based Tightly Coupled LiDAR-Inertial Odometry
Zelin Wang, Xu Liu, Limin Yang, Feng Gao
Abstract
This letter presents SW-LIO, a tightly coupled LiDAR-inertial odometry based on the sliding window approach. The proposed methodology encompasses rapid ground segmentation and the design of an iterative error-state Kalman filter (ESKF) to effectively fuse LiDAR point clouds and IMU measurements. By establishing a coupling relationship between the current state and the previous frames through the sliding window, the point clouds from the previous frames serve as a constraint for the current pose, resulting in more accurate state estimation. Furthermore, ground residual, bias residual and gravity residual are proposed, enabling more precise estimation of state variables beyond pose. These enhancements enable the system to deliver superior initial values for the filter in fast-moving or unstable environments, thereby improving the system's robustness. To evaluate the proposed framework, comprehensive testing has been conducted on public datasets as well as challenging real-world scenarios. The experimental results demonstrate that SW-LIO outperforms other state-of-the-art methods in terms of robustness and precision while maintaining similar time consumption.