HDA-LVIO: A High-Precision LiDAR-Visual-Inertial Odometry in Urban Environments with Hybrid Data Association

TL;DR

The paper addresses the challenge of precise localization in urban environments by introducing HDA-LVIO, a hybrid data association-based LiDAR-Visual-Inertial odometry. It jointly leverages LIS for ICP-based LiDAR-Inertial mapping and global-map alignment, and VIS for image-based plane projection, feature-depth estimation via a sliding window, and epipolar-based recovery, with reprojection errors fused through an Error State Iterated Kalman Filter. Key contributions include Incremental Adaptive Plane Extraction to maintain a stable plane-based projection regime, robust feature-depth recovery within sliding windows, and a unified fusion framework that exploits both projection and feature points. Extensive experiments on KITTI, NTU-VIRAL, and a custom platform demonstrate significant localization accuracy gains over multiple baselines, while highlighting real-time feasibility and potential computational/dynamic-object challenges for future work.

Abstract

To enhance localization accuracy in urban environments, an innovative LiDAR-Visual-Inertial odometry, named HDA-LVIO, is proposed by employing hybrid data association. The proposed HDA_LVIO system can be divided into two subsystems: the LiDAR-Inertial subsystem (LIS) and the Visual-Inertial subsystem (VIS). In the LIS, the LiDAR pointcloud is utilized to calculate the Iterative Closest Point (ICP) error, serving as the measurement value of Error State Iterated Kalman Filter (ESIKF) to construct the global map. In the VIS, an incremental method is firstly employed to adaptively extract planes from the global map. And the centroids of these planes are projected onto the image to obtain projection points. Then, feature points are extracted from the image and tracked along with projection points using Lucas-Kanade (LK) optical flow. Next, leveraging the vehicle states from previous intervals, sliding window optimization is performed to estimate the depth of feature points. Concurrently, a method based on epipolar geometric constraints is proposed to address tracking failures for feature points, which can improve the accuracy of depth estimation for feature points by ensuring sufficient parallax within the sliding window. Subsequently, the feature points and projection points are hybridly associated to construct reprojection error, serving as the measurement value of ESIKF to estimate vehicle states. Finally, the localization accuracy of the proposed HDA-LVIO is validated using public datasets and data from our equipment. The results demonstrate that the proposed algorithm achieves obviously improvement in localization accuracy compared to various existing algorithms.

Figures (18)

• Figure 1: The projection points and feature points utilized in this paper.
• Figure 2: The framework of the proposed HDA-LVIO.
• Figure 3: The flowchart of the proposed plane extraction method.
• Figure 4: Image frame utilized in feature point depth recovery.
• Figure 5: Feature point tracking failed due to rapid vertical movement.