Evaluation of Character Recognisers: Artificial Neural Network and Nearest Neighbour Approach

This paper is an aspect of artificial neural network that evaluates the accuracy of two character recognizers when applied to English Language characters. The two recognizers used are artificial neural network and Nearest Neighbour approach. The characters need to be represented in binary digits. To achieve this, the characters were embroiled into a grid made up seven column and five rows. Math-lab tool was used to implement the recognizer son the features extracted. The result shows Nearest Neighbour to be a better recognizer when applied to English Characters.