Classification of noisy English alphabets using neural network

This paper classifies noisy English Alphabets with the help of Artificial Neural Network. Here a supervised single layer perceptron learning algorithm for the classification of noisy English alphabets has been proposed. The presented algorithm requires very few number of input neurons for training. The algorithm is capable of classification of English alphabets for different scenario of noise. Simulation results show that algorithm is able to recognize 81.25% English alphabets in a very noisy environment having SNR 0.809.