Classification of Arabic script using multiple sources of information: State of the art and perspectives

Author First name, Last name, Institution

Najoua Essoukri Ben Amara
Faouzi Bouslama

Document Type

Article

Source of Publication

International Journal on Document Analysis and Recognition

Publication Date

7-1-2003

Abstract

Abstract.In this paper, we present a review of the state of the art in the current classification techniques used in the optical character recognition of the Arabic script (AOCR). We consider multiple sources of information-based hybrid approaches and multiple classifiers. We show that in order to improve classification results obtained with single classifiers, it is necessary to combine several sources of information either at the level of feature extraction/description, or at the classification stage, or at both levels. We provide a qualitative comparison and discuss the strengths and weaknesses of these approaches.

ISSN

1433-2825

Publisher

Springer Science and Business Media LLC

Volume

5

First Page

195

Last Page

212

Disciplines

Computer Sciences

Indexed in Scopus

no

Open Access

no

Share

COinS