Character Recogntion System: Performance Comparison of Neural Networks and Genetic Algorithm

In this paper, we present a Character Recognition System: Performance Comparison of Artificial Neural Networks and Genetic Algorithm for the character recognition using the Artificial Neural Networks (ANN) and Genetic Algorithm (GA) and measure performance by changing various selection criteria of both algorithms. Mainly, Back propagation Learning Neural Network Algorithm (BPN) as the ANN is used. This system has been taken the character´s image as its input. The input images have been filtered by filtering methods of image processing to remove noise and smoothing it and converted to binary image to detect its edges properly and clipped to get the actual image to input. Features of each individual clipped image have been extracted by taking a definite resize binary image value. These extracted features of input images of characters are used by the Back propagation Learning Neural Network Algorithm and Genetic Algorithm. Thus the network has been trained and creates a knowledge base of recognition. The same procedures have been applied for recognition but with only difference is that the neural network is used the previously learned weights and thresholds to calculate the output for BPN. In this paper, the performances of changing the various selection criteria of both algorithms have been also measured to learn and recognize the character.