On Line Recognition System for Arabic Handwritten Text

Abstract

This paper presents the design and the implementation of an On Line Arabic text Recognition system that is used for cursive handwritten recognition. In addition to Arabic characters, OLAR can recognize numerical characters, and special symbols. The direction and style of writing are used to compose the main components of the feature vector of the characters to be recognized. OLAR uses Euclidean distance approach and artificial neural networks for classification. The obtained results showed that OLAR can compete well with other handwriting recognition systems. The recognition rate ranges from 90% to 100%.