Neural network modeling and implementation of a handwriting system

Several mathematical models have been developed in the literature to characterize and study the handwriting process. But none of these methods have been implemented on any electronic devices. The work presented in this paper focuses on applied artificial neural networks for the modeling of the handwriting biological system. After studying the handwriting process, and based on Yasuhara's physical model, a neuronal model is proposed and then implemented on the Xilinx ZC702 embedded board. The results of the simulation and the implementation are applied to letters and parts of Arabic words. The obtained results are Satisfactory compared to Yasuhara's differential system responses.