2D autonomous robot localization using fast SLAM 2.0 and YOLO in long corridors
- Autonomous navigation is one of the main areas of research in mobile robots and intelligent connected vehicles. In this context, we are interested in presenting a general view on robotics, the progress of research, and advanced methods related to this field to improve autonomous robots’ localization. We seek to evaluate algorithms and techniques that give robots the ability to move safely and autonomously in a complex and dynamic environment. Under these constraints, we focused our work in the paper on a specific problem: to evaluate a simple, fast and light SLAM algorithm that can minimize localization errors. We presented and validated a FastSLAM 2.0 system combining scan matching and loop closure detection. To allow the robot to perceive the environment and detect objects, we have studied one of the best deep learning technique using convolutional neural networks (CNN). We validate our testing using the YOLOv3 algorithm.
Author of HS Reutlingen | Zimmermann, Alfred |
---|---|
Editor of HS Reutlingen | Zimmermann, Alfred |
DOI: | https://doi.org/10.1007/978-981-16-3264-8_19 |
Erschienen in: | Human centred intelligent systems : proceedings of KES-HCIS 2021 conference ; Smart innovation, systems and technologies, volume 244 |
Publisher: | Springer |
Place of publication: | Singapore |
Editor: | Alfred ZimmermannORCiD, Robert Howlett, Lakhmi Jain, Rainer Schmidt |
Document Type: | Conference proceeding |
Language: | English |
Publication year: | 2021 |
Tag: | SLAM; autonomous vehicles; convolutional neural networks; deep learning; lidar sensor |
Page Number: | 10 |
First Page: | 199 |
Last Page: | 208 |
DDC classes: | 004 Informatik |
Open access?: | Nein |
Licence (German): | In Copyright - Urheberrechtlich geschützt |