Related Experiment Video
Updated: Sep 21, 2025
11:53
The Modular Design and Production of an Intelligent Robot Based on a Closed-Loop Control Strategy
Published on: October 14, 2017
11.8K
Multi-LiDAR Mapping for Scene Segmentation in Indoor Environments for Mobile Robots
Pavel Gonzalez1, Alicia Mora1, Santiago Garrido1
1Robotics Lab, Universidad Carlos III de Madrid, 28911 Leganes, Spain.
Sensors (Basel, Switzerland)
|May 28, 2022
Summary
This study introduces multi-LiDAR sensor fusion for enhanced mobile robot mapping and navigation. Combining 2D and 3D data improves environmental perception and enables advanced features like room segmentation.
Area of Science:
- Robotics
- Computer Vision
- Sensor Fusion
Background:
- Current mobile robot applications primarily use 2D LiDAR for indoor tasks.
- Single data type mapping is insufficient for six-degree-of-freedom environments.
- Multi-LiDAR sensor fusion enhances robot mapping capabilities by integrating diverse data types.
Purpose of the Study:
- To develop advanced techniques for mapping and navigation in indoor environments using Multi-LiDAR sensor fusion.
- To improve robot perception and overcome limitations of single-sensor systems.
- To enable robust indoor localization and environmental understanding.
Main Methods:
- Implemented an Iterative Closest Point (ICP) scan matching algorithm with a distance threshold association counter as a multi-objective fitness function.
- Utilized Harmony Search for optimizing scan matching results without initial guesses or odometry.
- Developed a global Simultaneous Localization and Mapping (SLAM) approach to minimize accumulated errors.
- Integrated 2D and 3D mapping techniques, overlapping resulting maps for fused geometrical information at different heights.
- Proposed a room segmentation procedure analyzing fused geometrical data to overcome 2D map occlusions.
Main Results:
- Achieved superior mapping and navigation performance compared to solo odometry LiDAR matching.
- Successfully built global maps during SLAM, reducing accumulated errors.
- Demonstrated effective fusion of geometrical information from 2D and 3D maps.
- Validated the room segmentation procedure by implementing a successful door recognition system.
- Confirmed algorithm performance in both simulated and real-world experimental scenarios.
Conclusions:
- Multi-LiDAR sensor fusion significantly enhances indoor mobile robot mapping and navigation capabilities.
- The proposed ICP-based scan matching with Harmony Search optimization provides robust localization.
- Fused 2D and 3D mapping effectively addresses occlusions and improves environmental understanding.
- The developed room segmentation and door recognition systems showcase the practical benefits of the fusion approach.
