OCR Using Convolution Neural Network in Python with Keras and TensorFlow

Abstract

We aim to design an expert system for,” OCR using Neural Network” that can effectively recognize specific character of type style using the Artificial Neural Network Approach. We are pre-processing the input image, extracting the features, and then using the classification schema along with training of system to acknowledge the text. During this approach, we have trained the system to seek out the similarities, and also the differences among various handwritten samples. It takes the image of a hand transcription and converts it into a digital text. The extension of MNIST digits dataset has been used and A-Z characters in both uppercase and lowercase to detect handwritten text and convert it into digital form using Convolutional Neural Networks model, abbreviated as CNN, for text classification and detection also we are using keras graph to predict alphanumeric characters drawn using a finger and linked our handwriting text recognition program using keras and TensorFlow librar.