Character Recognition System Using Xilinx System Generator

Nowadays, bulk of printed documents are available in libraries, information centers, museums and offices which creates an increasing demand of Character Recognition. Character recognition is an area of research where many researchers have presented their work and it is still an area under research to achieve higher accuracy. This paper aims to present a technique for character recognition using edge detection. Initially the main goal is to carryout hardware implementation of edge detection technique using Xilinx System Generator. In this paper three different methods namely direct correlation with segmented templates, MATLAB edge detected image with edge detected templates and Xilinx System Generated Hardware co-simulated output image with Xilinx System Generated templates are compared. It is concluded that the accuracy of edge detection correlation technique is better than direct correlation technique. For edge detection Sobel operator is used. For hardware co-simulation Spartan 3E and Virtex 5 FPGA Boards are used.