Neural Networks are being used for character recognition from last many years but most of the work was confined to English character recognition. Till date a very little work has been reported for Handwritten Hindi Character recognition. In this paper, we have made an attempt to recognize handwritten Hindi characters by using a multilayer perceptron with one hidden layer. The error backpropagation algorithm has been used to train the MLP network. Also, an analysis has been carried out to determine the number of hidden nodes to achieve high performance of backpropagation network in the recognition of handwritten Hindi characters. The system has been trained using several different forms of handwriting provided by both male and female participants of different age groups. Finally, this rigorous training results an automatic HCR system using MLP network. In this work, the experiments were carried out on two hundred fifty samples of five writers. The results showed that the MLP networks trained by the error backpropagation algorithm is superior in recognition accuracy and memory usage. The result indicates that the backpropagation network provides good recognition accuracy of more than 80% of handwritten Hindi characters.We have also made an attempt to implement gradient feature extraction technique in order to recognize handwritten Hindi character recognition. A comparative analysis has been done regarding training time and recognition accuracy of handwritten characters by considering both global input and gradient feature input. It was found that gradient feature extraction technique has reduced training time drastically as compared to global input.

Keywords: Hindi character recognition(HCR), neural networks, multilayer perceptron (MLP) , error backpropagation network and gradient feature extraction technique.