LOGIN TO YOUR ACCOUNT

Username
Password
Remember Me
Or use your Academic/Social account:

CREATE AN ACCOUNT

Or use your Academic/Social account:

Congratulations!

You have just completed your registration at OpenAire.

Before you can login to the site, you will need to activate your account. An e-mail will be sent to you with the proper instructions.

Important!

Please note that this site is currently undergoing Beta testing.
Any new content you create is not guaranteed to be present to the final version of the site upon release.

Thank you for your patience,
OpenAire Dev Team.

Close This Message

CREATE AN ACCOUNT

Name:
Username:
Password:
Verify Password:
E-mail:
Verify E-mail:
*All Fields Are Required.
Please Verify You Are Human:
fbtwitterlinkedinvimeoflicker grey 14rssslideshare1
Languages: French
Types: Article
Subjects: Analyse d'images, Vision par ordinateur, Apprentissage, Classification supervisée, Treillis de Galois, Treillis de concepts, Robotique mobile autonome, Localisation topologique, Image analysis, Computer vision, Learning, Supervized classification, Galois lattices, Concept lattices, Autonomous mobile robotics, Topological localization, 629.8
Cette thèse s'inscrit dans le cadre de la classification d'images supervisée appliquée à la robotique mobile autonome en milieu structuré. Pour naviguer, et en particulier pour se repérer, un robot utilise des amers attachés à des sites de l'environnement, modélisé par un graphe topologique. À chaque noeud du graphe est associé un amer. Les amers sont des combinaisons d'attributs visuels optimales, obtenues au travers d'un formalisme particulier appelé "treillis de Galois" ou "treillis de concepts". Des algorithmes de construction de treillis ont été modifiés afin de permettre, de façon incrémentale, l'établissement d'une classification supervisée des images et l'extraction des amers visuels. Une extension probabiliste et une approche locale ont été implémentées pour améliorer les performances de l'apprentissage. Enfin, une application robotique complète a été implémentée dans les laboratoires du LAAS-CNRS et de SUPAERO. The subject if this Thesis is situated within the field of image supervized classification applied to autonomous mobile robotics in a structural environment. To navigate, and more especially to locate itself, a robot needs landmarks, attached to each site of the environment, which is represented by a topological graph. The robot attaches a landmark to each node of the topological graph. Landmarks are combinations of features obtained through a specific formalism called "Galois lattices" or "concept lattices". Some lattice building algorithms are modified to allow incremental supervized image classification to extract visual landmarks. A probabilistic expansion and a local approach have been developed to perform our system. Finally, a robotics application is described and has been implemented in the LAAS-CNRS and SUPAERO laboratories.
  • No references.
  • No related research data.
  • No similar publications.

Share - Bookmark

Cite this article