Polymorphic Circuit Design for Speedy Handwritten Character Recognition Using Cartesian Genetic Programming

Abstract

In this paper we have used CGP (Cartesian Genetic Programming) to obtain an optimum polymorphic circuit for speedy recognition of hand written characters. Images are converted to binary form and arranged in one-dimension array. This data is fed, in parallel, to the system. Experiments have demonstrated that promising results can be obtained using a much simpler circuit having 1-bit multiplexers. All the experiments are conducted without any pre-processing and image is being fed in parallel to the system, thus reducing complexity of the circuit and time for recognition enormously.