Enhancing Multi-Robot SLAM: Centralized LiDAR-Based Loop Closure Detection Approach

Authors

Basma Ahmed Jalil, Department of Computer Engineering, Faculty of Engineering, Mosul University, Mosul 41002, IraqFollow
Ibraheem Kasim Ibraheem, Department of Electrical Engineering, College of Engineering, University of Baghdad, Baghdad 10001, Iraq AND Department of Electronics and Communication Engineering, Uruk University, Baghdad, IraqFollow

Abstract

The loop closure detection is crucial for global mapping and route correction in multi-robot simultaneous localization and mapping (SLAM). However, including loop closure detection algorithms in MR-SLAM increases the computational complexity and the required resources on the robot board and at the base station. In this paper, An Enhanced Multi-Robot Fast Localization Odometry and Mapping (EMR-FLOAM) to deal with computation complexity issue. The EMR-FLOAM algorithm addresses computational complexity and resource requirements by utilizing a two-stage non-iterative distortion compensation technique, resulting in optimized code and accelerated localization and map construction processes. The simulation of the proposed work has been tested on the outdoor dataset tailored for multi-robot systems and the indoor environment constructed by the Gazebo simulator. The simulation results have been compared with Enhanced Multi Robot An improved Localization odometry and Mapping (EMR-ALOAM) algorithm and it was noted that EMR-FLOAM shows an enhancement in reducing the error drift, but the error is still larger than that of EMR-ALOAM, while the computational complexity in EMR-FLOAM is smaller than that of EMR-ALOAM.

Recommended Citation

Jalil, Basma Ahmed and Ibraheem, Ibraheem Kasim (2024) "Enhancing Multi-Robot SLAM: Centralized LiDAR-Based Loop Closure Detection Approach," Iraqi Journal for Computer Science and Mathematics: Vol. 5: Iss. 4, Article 24.
DOI: https://doi.org/10.52866/2788-7421.1219
Available at: https://ijcsm.researchcommons.org/ijcsm/vol5/iss4/24