Handwritten Character Recognition Application by Using Cellular Neural Network

Abstract

Hand-written character recognition is one of the important fields of pattern recognition. Within the scope of this area of important documents and archives and other written texts transfering to digital media or recognition of the printer tries to unravel the problems. Many algorithms have been developed for these problems. Algorithms that have been developed to be desired, the high accuracy rate and being applicable for numeric desings like FPGA. Therefore, for classification, feature vector is extracted by using Gabor-like Cellular Neural Network (HSA) filters. These filters are implemented with efficient algorithms on FPGA [10]. By this means, an algorithm has been developed FIR filters designed by the Gabor more efficient in terms of processing time and accuracy, the percentage of capital letters, which at around 80%.