Summary
International Workshop on Smart Info-Media Systems in Asia
2023
Session Number:RS4
Session:
Number:RS4-1
A Vectorized Map Library as a Platform for Higher Layer Planning for Mobile Robots
Wataru Mita, Johei Matsuoka, Kazuya Tago,
pp.122-127
Publication Date:2023/8/31
Online ISSN:2188-5079
DOI:10.34385/proc.77.RS4-1
PDF download
Summary:
Simultaneous localization and mapping (SLAM) outputs the location information of objects in the environment as a point cloud map in a raster format. For extensions to existing SLAM functions, such as semantic SLAM, it is difficult to determine whether the map representation is suitable for adding semantic information. In some cases, it is difficult to use this configuration. In this study, we propose a method to vectorize the raster map generated by SLAM. This vector map realizes a platform for higher-layer planning for mobile robots. Specifically, we implement the library which includes noise reduction, missing interpolation for raster maps, typing for environmental components, shape extraction, vectorization, and other spatial segmentation functions. This library unifies methods that previously differed from developer to developer. It allows planning to be implemented using geometric calculations without the need for specialized knowledge.