Optical Character Recognition and Classification using Machine Learning for Printed Images

  • Mr. Prashant Sopanrao Kolhe, Dr. Ulhas Shiurkar

Abstract

Nowadays, there are numerous innovative methodologies claimed toward the expanding requirements in newly developing operations; with specific methods, there are several procedures already done by various researchers the text identification of reproduction in various native languages. Additionally, there is a lot of activity prepared for the composed corporeality, but it is only restricted to the laboratory. However, it has not been utilized effectively. So in this article, introduced the smallest interval classifier scheme for OCR System of indented as well as scanned newsprint Marathi scenario. In this work, we proposed an OCR character recognition and classification using machine learning approach. The printed document image dataset has used to extract text information from image and validate with ground truth text information. In experiment analysis accuracy has evaluated of extracted text, that demonstrates around 95% accuracy over the heterogeneous text data..

Published
2019-03-31
How to Cite
Mr. Prashant Sopanrao Kolhe, Dr. Ulhas Shiurkar. (2019). Optical Character Recognition and Classification using Machine Learning for Printed Images. International Journal of Advanced Science and Technology, 29(3), 15596-15603. Retrieved from http://sersc.org/journals/index.php/IJAST/article/view/35678
Section
Articles