Hermite and Gabor transforms for noise reduction and handwriting classification in ancient manuscripts

Abstract : In this paper, we propose a biologically inspired, global segmentation free methodology for manuscript noise reduction and classification. Our method consists of developing well adapted tools for writing enhancement, background noise, text and drawing separation and handwritten patterns characterization with orientation features. We have used here the spectral domain of handwritten images by frequency decompositions (Hermite transforms) and Gabor filtering for selective text information extraction. The method which has been developed for manuscript authentication has been tested on the 18th century Montesquieu’s corpus. The current results are very promising and show that it is possible to analyze handwritten drawings without any a priori grapheme segmentation.
Type de document :
Article dans une revue
IJDAR, 2007, 2-4, 9, pp.101-122
Liste complète des métadonnées

https://hal.archives-ouvertes.fr/hal-01535448
Contributeur : Équipe Gestionnaire Des Publications Si Liris <>
Soumis le : jeudi 8 juin 2017 - 23:30:24
Dernière modification le : vendredi 10 novembre 2017 - 01:20:19

Identifiants

  • HAL Id : hal-01535448, version 1

Collections

Citation

Véronique Eglin, Stéphane Bres, Carlos Rivero-Moreno. Hermite and Gabor transforms for noise reduction and handwriting classification in ancient manuscripts. IJDAR, 2007, 2-4, 9, pp.101-122. 〈hal-01535448〉

Partager

Métriques

Consultations de la notice

96