HCTO: Optimality-Aware LiDAR Inertial Odometry with Hybrid Continuous Time Optimization for Compact Wearable Mapping System

TL;DR

A novel LIO for WMSs based on Hybrid Continuous Time Optimization (HCTO) considering the optimality of Lidar correspondences is proposed considering the optimality of Lidar correspondences to achieve real-time performance and better odometry accuracy.

Abstract

Compact wearable mapping system (WMS) has gained significant attention due to their convenience in various applications. Specifically, it provides an efficient way to collect prior maps for 3D structure inspection and robot-based "last-mile delivery" in complex environments. However, vibrations in human motion and the uneven distribution of point cloud features in complex environments often lead to rapid drift, which is a prevalent issue when applying existing LiDAR Inertial Odometry (LIO) methods on low-cost WMS. To address these limitations, we propose a novel LIO for WMSs based on Hybrid Continuous Time Optimization (HCTO) considering the optimality of Lidar correspondences. First, HCTO recognizes patterns in human motion (high-frequency part, low-frequency part, and constant velocity part) by analyzing raw IMU measurements. Second, HCTO constructs hybrid IMU factors according to different motion states, which enables robust and accurate estimation against vibration-induced noise in the IMU measurements. Third, the best point correspondences are selected using optimal design to achieve real-time performance and better odometry accuracy. We conduct experiments on head-mounted WMS datasets to evaluate the performance of our system, demonstrating significant advantages over state-of-the-art methods. Video recordings of experiments can be found on the project page of HCTO: \href{https://github.com/kafeiyin00/HCTO}{https://github.com/kafeiyin00/HCTO}.

Figures (17)

• Figure 1: Hardware configuration of the compact helmet-based wearable sensing system. The main sensor is the Livox MID360.
• Figure 2: Prior map constructed by Hybrid Continuous Time Optimization (HCTO) for robot navigation in different challenging scenes using wearable devices. (a) The mapping results from Fast-Lio2 xu2022fast are corrupted by long-term drift, where the building walls are slanted and the ground plane bends downwards. The robot goes to the wrong place using the map. (b) The mapping results using the proposed HCTO maintain good accuracy, where the building walls are vertical and the ground plane maintains levelness. The robot goes to the right goal.
• Figure 3: Maximum-A-Priori (MAP) optimization in the active window taking the $4^{th}$ order B-Spline as an example.
• Figure 4: System overview of the proposed helmet-based LIO system. (a) LiDAR Inertial Odometry (LIO) with Hybrid Continuous Time Optimization (HCTO). (b) Diagram of HCTO.
• Figure 5: Illustration of different motion states in the repetitive motion pattern.