This paper presents a new algorithm for fast mobile robot self-localization in structured indoor environments based on geometrical and analytical matching, GEMA2. The proposed method takes advantage of the available structural information to perform a geometrical matching with the environment information provided by measurements collected by a laser rangefinder. In contrast to other global self-localization algorithms like Monte Carlo or SLAM, GEMA2 provides a linear cost with respect the number of measures collected, making it suitable for resource-constrained embedded systems. The proposed approach has been implemented and tested in a mobile robot with limited computational resources showing a fast converge from global self-localization.
SANCHEZ BELENGUER Carlos;
SORIANO Angel;
VALLES Marina;
VENDRELL VIDAL Eduardo;
VALLERA Angel;
2014-06-13
ELSEVIER SCIENCE BV
JRC88641
0921-8890,
http://www.sciencedirect.com/science/article/pii/S0921889014000128,
https://publications.jrc.ec.europa.eu/repository/handle/JRC88641,
10.1016/j.robot.2014.01.009,
| Name | Country | City | Type |
|---|
This document is only visible at the Commission level.
You are not authorized to publish or distribute it outside the European Commission.
This is a public document. You can share this publication.
Datasets
| ID | Title | Public URL |
|---|
Dataset collections
| ID | Acronym | Title | Public URL |
|---|
Scripts / source codes
| Description | Public URL |
|---|
Additional supporting files
| File name | Description | File type |
|---|