Improvements to the FLOAM Algorithm: GICP Registration and SOR Filtering in Mobile Robots with Pure Laser Configuration and Enhanced SLAM Performance

Read full text Summarise with VIRAT
Authors: Shichen Fu, Tianbao Zhao, Junkai Zhang, Guangming Guo, Weixiong Zheng
Cited by: 0
References: 34

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Laser SLAM is a key enabling technology for autonomous navigation of intelligent mobile robots. The standard FLOAM algorithm experiences low positioning accuracy, weak anti-interference performance, and prone error accumulation in pure LiDAR scenarios, making it difficult to meet practical engineering requirements. The focus of numerous studies is thus on improved pure laser SLAM algorithms that are highly robust. The enhanced algorithm of FLOAM GICP registration and SOR filtering is applied in this study. The SOR filtering processes the laser point cloud to remove outlier noise. The GICP registration replaces the classic with an optimized matching cost function. Experiments are conducted on a mobile robot with a Leishen C16 LiDAR to simulate real-life tests in an indoor corridor and outdoor plaza on the Gazebo simulation platform. The results from the EVO tool’s quantitative evaluation indicate that the indoor mean absolute error and RMSE were reduced by 46.67% and 41.67% compared with FLOAM. The outdoor mean and maximum errors are reduced by 46.00% and 70.00%, respectively. The proposed improved scheme achieves centimeter-level positioning accuracy and strong robustness in pure laser configurations without auxiliary sensors such as IMUs or odometers, providing a reliable technical solution for the engineering application of mobile robots in sensor-constrained scenarios.

Fields of Study & Subjects

Robotics and Sensor-Based Localization 3D Surveying and Cultural Heritage Inertial Sensor and Navigation

Publication details

Journal / source
Applied Sciences
Volume
16
Issue
7
Pages
3141–3141
Publication date
24 March 2026
Language
EN