Information Technology Reference
In-Depth Information
6
Conclusions and Future Work
In this article, a novel way to extract topological maps and connectivity information
from standard grid-like grayscale representations is described. The approach presented
gains advantage from its simplicity, accuracy and performance. One possible disad-
vantage of using the EVG-THIN method to compute the skeleton of the grid map is
the dependency on the correct parameterization, which is not straightforward for most
cases. However, the approach presented in this paper is not limited to the use of EVG-
THIN to extract the skeleton, other techniques like those mentioned in Section 2, can
also be used.
Unlike most previous works in this area, here the intent is not to present solely a rep-
resentation of the graph on top of the grid, but also to give one step ahead by proposing
a way to convert visual information into data structures, by means of image processing
techniques as described.
The proposed approach offers, as output, a complete characterization of the topolog-
ical aspects of the environment, which has the ability to assist robot's navigation in a
broad spectrum of activities, especially those that include path planning. This technique
has been recently used to provide a topological map for mobile robots in cooperative
patrolling misions [21]
As for future work, it would be interesting to test this approach using different meth-
ods in the literature to obtain the underlying diagram to check whether it is possible to
speed up the first step of the algorithm without losing quality on the topological repre-
sentation. Additionally, some questions are still left open like addressing fast update of
the Voronoi Diagram given dynamic changes in the environment as well as consider-
ing 3D models and deal with topological navigation using mobile robots in real world
scenarios.
Acknowledgements.
This work was financially supported by a PhD grant (SFRH/BD/
64426/2009) from the Portuguese Foundation for Science and Technology (FCT) and
the Institute of Systems and Robotics (ISR-Coimbra) also under regular funding by
FCT.
References
1. Thrun, S.: Learning maps for indoor mobile robot navigation. Artificial Intelligence 99,
21-71 (1998)
2. Zimmer, U., Fischer, C., Von Puttkamer, E.: Navigation on topologic feature-maps. In: 3rd
International Conference on Fuzzy Logic, Neural Nets and Soft Computing, Fukuoka, Japan,
pp. 131-132 (1994)
3. Dudek, G., Jenkin, M., Milios, E., Wilkes, D.: Map validation and robot self-location in a
graph-like world. Robotics and Autonomous Systems 22(2), 159-178 (1997)
4. Dudek, G., Jenkin, M., Milios, E., Wilkes, D.: Topological exploration with multiple robots.
In: 7th International Symposium on Robotics with Applications (ISORA 1998), Anchorage,
Alaska, May 10-14 (1998)
5. Ferreira, F., Davim, L., Rocha, R., Dias, J., Santos, V.: Presenting a technique for registering
images and range data using a topological representation of a path within an environment. J.
of Automation, Mobile Robotics & Intelligent Systems 1(3), 47-55 (2007)
