Pen-based recognition system of Arabic character utilizing structural and fuzzy techniques

Abstract

Describes a hybrid approach for the automatic recognition of handwritten Arabic characters. The algorithm is based on structural techniques and fuzzy logic. Local features such as lines, curves, diacritic points are extracted from the geometry and topology of characters. Fuzzy linguistic variables are used to model the features and provide a suitable mean to vaguely describe the many styles and variations in the writing system. The combination of local features provide a natural way to describe characters in a compact style. Fuzzy if-then rules are used for classification. This hybrid technique is efficient for large and complex sets such as Arabic characters.