Deep Learning Technique for Recognizing Hand Written Characters

Abstract

Recognizing character is one of the main task for a human. In our proposed model we present a method that which recognizes the characters which are handwritten. This process was performed using deep neural networks. The process of training the considered dataset with neural networks has become easier. Where the GPU availability also made in training improvement using the neural networks. On other side some of the services like Google cloud and Amazon web services are also providing the cloud platforms to train the neural networks. Where, it become easier and available to train using the deep learning algorithms using the cloud services. Here, we designed a system that which recognizes the hand written characters based on the image segmentation. To recognize the characters we have used CNN method of the deep learning for the image processing and for the training of network we used TensorFlow. Emnist dataset is considered for the training and testing the architecture.