Faculte des Sciences de l'Ingenieur, Departement d'Electronique, Universite de Batna, Algerie.
In this paper we propose a new technique of handwritten Arabic characters recognition based on the use of neural network. Exploiting features fluently used and proceeding to a new programming manner of these networks. Our system presents a well better aptitude to recognize those characters.
Optical character recognition, Invariant geometrical moments, Artificial neural network, pattern recognition, feature extraction
