A Comparative Analysis on Bangla Handwritten Digit Recognition with Data Augmentation and Non-Augmentation Process

Abstract

Determination of Bangla handwritten digit is a momentous image classification task. Though object recognition technology is getting smarter day by day, still Bangla handwritten digit recognition remains inconclusive. Researchers are becoming more concerned about handwritten digit recognition for it’s educational and advantageous importance. But it is a matter of trouble that the improvement in Bangla handwritten digit recognition is significantly less as compared to the other languages. To improve the performance of the Bangla handwritten digit recognition system, we have designed a model, in which all basic Bangla digits have been classified. Furthermore, we have also demonstrated Densenet121 architecture in our system. For recognizing Bangla handwriting digits, we proposed CNN (Convolution Neural Network) model. Our system has been experimented on the NumtaDB dataset for recognizing Bangla digit both with augmentation and non-augmentation.