Handwritten Text Recognition Using CNN

Handwritten Text Recognition Using CNN

Author Name : Rishabh Goel, Ajay Kaushik

ABSTRACT

Recognition  of  handwritten  characters  is  one  of  the  most  thought-provoking  areas  of  pattern  recognition. It contributes enormously to the advancement of automation process and improves the interface between man & machine in numerous applications. It  is  useful  while dealing  with  practical  problems,  signature  verification,  interpretation  of postal address, mailing  bank  check processing, documentation analysis, also document verification and many  others.

The goal of this project is to create a model that will be able to recognize and determine the handwritten characters from its image with better accuracy and convert it to Digital text Format. The concepts of Convolution Neural Network (CNN) with various architectures are used to train a model on the IAM Handwritten dataset that can accurately classify words. The major goal of the project is understanding of Convolutional Neural Network, and applying it to the Handwritten Text Recognition system.

Keywords: Handwritten Text Recognition, Neural Network, CNN, Deep Learning, HTR