Skip to Main content Skip to Navigation

Appariement inexact de graphes appliqué à la recherche d'image et d'objet 3D.

Abstract : Graphs are models of representation that can model many type of documents. In this thesis, we focus on their use for research in multimedia databases. We begin by presenting the theory of graphs and around an overview of methods that have been proposed for matching. Then, we are particularly interested in their use for recognition forms and multimedia indexing. In order to respond in the most generic possible different research problems, we propose to work within the framework of kernel functions. This framework allows to separate the problems related to the nature of the documents those introduced by the di fferent types of research. Thus, all our energy is devoted to the design of mapping functions, but bearing in mind that they must meet a number mathematical properties. In this context, we propose new solutions that better meet the speci c graphs from primitive and visual descriptors. We also present algorithms to quickly assess these functions. Finally, we present experiments that highlight these different characteristics and experiences that show advantages of our models with respect to the literature.
Complete list of metadatas

Cited literature [87 references]  Display  Hide  Download
Contributor : Michel Jordan <>
Submitted on : Tuesday, November 22, 2011 - 10:57:40 AM
Last modification on : Thursday, December 17, 2020 - 12:14:37 PM
Long-term archiving on: : Thursday, February 23, 2012 - 2:21:46 AM


  • HAL Id : tel-00643534, version 1


Justine Lebrun. Appariement inexact de graphes appliqué à la recherche d'image et d'objet 3D.. Traitement du signal et de l'image [eess.SP]. Université de Cergy Pontoise, 2011. Français. ⟨tel-00643534⟩



Record views


Files downloads