Handwritten Multiscript Numeral Recognition using Artificial Neural Networks
Stuti Asthana1, Farha Haneef2, Rakesh K Bhujade3
1Stuti Asthana is M.Tech Scholar from RKDF Institute of Science and Technology, Bhopal, India.
2Farha Haneef is M.Tech Coordinator in RKDF Institute of Science and Technology, Bhopal, India.
3Rakesh K Bhujade is with Technocrats Institute of Technology, Bhopal , India.
Manuscript received on February 20, 2011. | Revised Manuscript received on February 27, 2011. | Manuscript published on March 05, 2011. | PP: 1-5 | Volume-1 Issue-1, March 2011. | Retrieval Number: A001021111
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Published By: Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Handwritten Numeral recognition plays a vital role in postal automation services especially in countries like India where multiple languages and scripts are used .Because of intermixing of these languages; it is very difficult to understand the script in which the pin code is written. Objective of this paper is to resolve this problem through Multilayer feed-forward back-propagation algorithm using two hidden layer. This work has been tested on five different popular Indian scripts namely Devnagri, English, Urdu, Tamil and Telugu. Network was trained to learn its behavior by adjusting the connection strengths on every iteration. The resultant of each presented training pattern was calculated to identify the minima on the error surface for each training pattern. Experiments were performed on samples by using two hidden layers and the results revealed that as the number of hidden layers is increased, more accuracy is achieved in large number of epochs.
Keywords: Numeral Recognition, Artificial Neural Network, Supervised learning, Back Propagation Algorithm.