The handprinted of character recognition using Gabor filter

摘要

This paper proposes a new approach for handwritten character recognition, using a Ganor filter. As Gabor filters can be implemented using Cellular Neural Networks, this method is expected to lead the way to a very fast character recognition system. Gabor filter detect the direction of edges, and the directions are added. The character is recognised by this function called “orientation map". Then this function does not influence the position of a character, size and thickness.