| Full text | |
| Author(s): |
Total Authors: 2
|
| Affiliation: | [1] University of São Paulo. Institute of Mathematics and Computer Science. Department Computer Systems - Brasil
[2] University of Southern California. Department of Computer Science. Robotic Embedded Systems Laboratory - Estados Unidos
Total Affiliations: 2
|
| Document type: | Journal article |
| Source: | Journal of the Brazilian Computer Society; v. 13, n. 4, p. 69-80, 2007-12-00. |
| Abstract | |
Mapping is a basic capability for mobile robots. Most applications demand some level of knowledge about the environment to be accomplished. Most mapping approaches in the literature are designed to perform in small structured (indoor) environments. This paper addresses the problems of localization and mapping in large urban (outdoor) environments. Due to their complexity, lack of structure and dimensions, urban environments presents several difficulties for the mapping task. Our approach has been extensively tested and validated in realistic situations. Our experimental results include maps of several city blocks and a performance analysis of the algorithms proposed. (AU) | |
| FAPESP's process: | 05/04378-1 - Development of estimation and learning algorithms for embedded systems in mobile robots |
| Grantee: | Denis Fernando Wolf |
| Support Opportunities: | Scholarships in Brazil - Post-Doctoral |