Content-Based Old Documents Indexing

Mickaël Coustaty (1) , N. Sidère (2) , Jean-Marc Ogier (1) , Pierre Héroux (2) , Jean-Yves Ramel (3) , Chouaib Hassan (4) , Nicole Vincent (4) , Salim Jouili (5) , Salvatore Tabbone (5)

1 L3I - Laboratoire Informatique, Image et Interaction - EA 2118
2 LITIS - Laboratoire d'Informatique, de Traitement de l'Information et des Systèmes
3 LIFAT - Laboratoire d'Informatique Fondamentale et Appliquée de Tours
4 CRIP5 - EA 2517 - Centre de Recherche en Informatique de Paris 5
5 QGAR - Querying Graphics through Analysis and Recognition

A huge amount of printed documents were published and distributed during the 15th century. In order to protect this inheritance, a digitalization campaign is held on these documents. The mass of documents generated by digitization create a problem to retrieve and to index them. This paper presents a french collaborative project that brings together seven laboratories around ancient documents indexing by content. This project is composed of two steps: 1) Extraction of information from images and 2) Retrieving documents with similar semantic interpretation for user. Two actors of this project present in this paper how they combine their works and their goal.

Keywords: Old documents; CBIR; Indexing; Graph

Domains: Computer Vision and Pattern Recognition [cs.CV]

Mickaël Coustaty, N. Sidère, Jean-Marc Ogier, Pierre Héroux, Jean-Yves Ramel, et al.. Content-Based Old Documents Indexing. Eight International Workshop on Graphics Recognition - GREC 2009, Jul 2009, La Rochelle, France. pp.217-223. ⟨hal-00382086⟩