Publications / 2020 Proceedings of the 37th ISARC, Kitakyushu, Japan

Evaluating SLAM 2D and 3D Mappings of Indoor Structures

Yoshihiro Nitta, Derbew Yenet Bogale, Yorimasa Kuba and Zhang Tian
Pages 821-828 (2020 Proceedings of the 37th ISARC, Kitakyushu, Japan, ISBN 978-952-94-3634-7, ISSN 2413-5844)

This paper introduces a navigation algorithm of mobile indoor unmanned ground vehicle (UGV). The navigation methodology with AR markers is presented and demonstrated in detail. In the navigation algorithm, the mobile indoor UGV can make 2D or 3D map inside the structure. From the driving test, it has seen that the navigation algorithm with AR marker is essential to control the attitude of the UGV and drive autonomously. The proposed navigation algorithm is very important to drive UGV autonomously inside building. Lastly, for investigating the capability of Simultaneous Localization and Mapping (SLAM) data, the 2D and 3D maps are evaluated by comparing to traditional survey and structure from motion (SfM). In conducting the map, slowing the speed of UGV affects the 2D map negatively, while it has positive impact in 3D mapping. Using visual SLAM with LiDAR makes 3D map very easily and rapidly as compared to SfM. From these results, the proposed navigation algorithm and manufactured prototype UGV with the mapping device for 2D and 3D are useful for studying the inside buildings even in the developing countries

Keywords: SLAM; Visual SLAM; LiDAR; UGV; AR Marker