A Study on Deep Learning Approach for Handwritng Text Recognition in Images

This paper discusses on Handwriting text recognition that extract text from an image which converts it into digital form. The need for data digitization is increased in recent years. Handwriting recognition is an ongoing field of research. The high variance in style of writing, size, orientation, alignment, low brightness of text in images is a challenging issue. This study aims on deep learning techniques which are applied to handwriting text images for improved accuracy and recognition levels. Deep learning recognize object in an image using Convolutional Neural Network (CNN). Additionally, this paper focuses on preprocessing, text segmentation and feature extraction using CNN for text recognition of images into digital form.