Drop Caps Decomposition For Indexing - A New Letter Extraction Method

Mickaël Coustaty (1) , Jean-Marc Ogier (1) , Rudolf Pareti (2) , Nicole Vincent (2)

1 L3I - Laboratoire Informatique, Image et Interaction - EA 2118
2 CRIP5 - EA 2517 - Centre de Recherche en Informatique de Paris 5

This paper presents a new method to extract shapes in drop caps and particularly the most important shape: Letter itself. This method relies on a combination of a Aujol and Chambolle algorithm threaded with a segmentation using a Zipf Law in a second step. This method can be enhanced as a three-step process: 1)Decomposition in layers 2)Segmentation using a Zipf Law 3)Selection of connected components to only emphasize the required information- letter itself.

Mots clés: Document; Drop Caps; Letter; Extraction; Indexing

Domaines: Traitement du texte et du document

Mickaël Coustaty, Jean-Marc Ogier, Rudolf Pareti, Nicole Vincent. Drop Caps Decomposition For Indexing - A New Letter Extraction Method. 10th International Conference on Document Analysis and Recognition, Jul 2009, Barcelona, Spain. pp.476-480, ⟨10.1109/ICDAR.2009.233⟩. ⟨hal-00435374⟩