Online Robust Sliding-Windowed LiDAR SLAM in Natural Environments

TL;DR

This paper introduces an online graph-based 2D LiDAR SLAM system tailored for natural environments, combining robust weighting, sliding-window optimization, and parallel computing to achieve real-time, stable performance in cluttered habitats.

Contribution

It presents a novel real-time SLAM system specifically designed for natural habitats, integrating multiple techniques for robustness and efficiency.

Findings

System achieves stable performance in cluttered natural environments.

Real-time operation demonstrated through simulated and experimental results.

Efficient design confirms feasibility for autonomous environmental monitoring.

Abstract

Despite the growing interest for autonomous environmental monitoring, effective SLAM realization in native habitats remains largely unsolved. In this paper, we fill this gap by presenting a novel online graph-based SLAM system for 2D LiDAR sensor in natural environments. By taking advantage of robust weighting scheme, sliding-windowed optimization, fast scan-matcher and parallel computing, our system not only delivers stable performance in cluttered surroudings but also meets real-time constraint. Simulated and experimental results confirm the feasibility and efficiency in the overall design of the proposed system.