Abstract: The handwritten alphabet recognition based on  LVQ neural network is proposed.The principle of  LVQ neural network is introduced and applied in the complex alphabet recognition because of its simple structure,good generalization,and rapid convergence.Through preprocessing the alphabet  images and feature abstracting and training the network by the abstracted feature,and  the alphabet is recognized  by the trained network.The Matlab simulation results shows that the rate of alphabet recognition is high.

Key words: LVQ neural network, alphabet recognition, preprocess, feature abstract