Handwritten Digit and Text Recognition Based On Convolutional Neural Network Approach

The goal of this undertaking is to foster a powerful penmanship acknowledgment methods utilizing ideas of Machine learning and PC vision. An expansion of MNIST digits dataset called the Emnist dataset has been utilized. It contains 62 classes with 0-9 digits and A-Z characters in both capitalized and lowercase. To recognize transcribed content and convert it into computerized structure utilizing Convolutional Neural Network and Support Vector Machine, shortened as CNN and SVM, for text arrangement and identification, has been made. Before that we pre-prepared the dataset and applied different channels over it. Our framework will perceive the content precisely utilizing tensorflow libraries.