Online Arabic/Persian character recognition using neural network classifier and DCT features

Online handwriting recognition is gaining interest due to the increase of pen computing applications and availability of tablet devices. The recognition of Arabic/Persian (A/P) characters is different from western handwriting, in which different calligraphic styles and cursive nature makes automatic recognition a more challenging and complicated task. In this paper, a new method is proposed to represent A/P characters. The proposed method incorporates a new set feature vectors suitable for A/P character set. A recognition system utilizing these set of features is developed for handwritten A/P characters. The result of the overall recognition system compare favorably with previous techniques.