Abstract: In this paper, a novel technique to Arabic character recognition is presented. The system is based on the classified vector quantization (CVQ) technique. To prove the superiority of the proposed system, its performance is compared to the state-of-the-art artificial neural network (ANN) approach. The proposed system uses the minimum Euclidean distance to classify the input character. Simulation results indicate that the proposed system always outperforms the ANN system.

Keywords: Vector quantization, Minimum-distance classifier, ANN, MSE, standard deviation.