|1/2013 - 11|
An Effect of Noise in Printed Character Recognition System Using Neural NetworkGHEORGHITA, S. , MUNTEANU, R. , GRAUR, A.
|Click to see author's profile on SCOPUS, IEEE Xplore, Web of Science|
|Download PDF (706 KB) | Citation | Downloads: 412 | Views: 1,979|
backpropagation, character recognition, neural networks, noise perturbation, training algorithm
neural(16), networks(9), recognition(8), network(5), character(4)
Blue keywords are present in both the references section and the paper title.
About this article
Date of Publication: 2013-02-28
Volume 13, Issue 1, Year 2013, On page(s): 65 - 68
ISSN: 1582-7445, e-ISSN: 1844-7600
Digital Object Identifier: 10.4316/AECE.2013.01011
Web of Science Accession Number: 000315768300011
SCOPUS ID: 84875336154
In this article we present the implementation of a neural network model trained with a high noise level using a backpropagation algorithm and the experimental results for printed character recognition, based on the idea of using the primary information by reorganising it in a different format. The values obtained at the outputs of each network are processed by using analysis algorithms designed for this purpose. The suggested model is made up of two neural networks and two analysis modules. In M1 Module we designed a value analysis algorithm for all the outputs of the two neural networks in order to select the best values provided by the networks. The M2 Module also contains a designed algorithm, which assesses the data based on the fact that the highest values are directly correlated with the probability of correctly identifying the characters entered into the networks. Results are obtained for noise of up to 50% applied to the input data. The values obtained at the outputs of the two modules emphasises the increase of the printed character recognition level up to 89.1% for the M1 module and up to 89.8% for the M2 module, the number of errors decreasing vis-a-vis the RNA2 network response from 12.5% to 10.9%, and 10.2%, respectively. In order to set up the hidden layer of 90 neurons, a value of 92% was obtained at the output of the M2 analysis module.The performed model increased the printed character recognition rate by using the same primary information in a different manner. The validity and functionality of the suggested model are confirmed by experimental results.
|References|||||Cited By «-- Click to see who has cited this paper|
| G. L. Martin, J. A. Pittman, "Recognizing hand-printed letters and digits using backpropagation learning", Neural Computation, vol. 3, no. 2, pp. 258-267, Summer 1991. |
 M. Fukumi, S. Omatu, F. Takeda, T. Kosaka, "Rotation-invariant neural pattern recognition system with application to coin recognition", IEEE Trans. Neural Networks, vol. 3, no. 2, 1992.
[CrossRef] [Web of Science Times Cited 82] [SCOPUS Times Cited 91]
 Z. Saidane, C. Garcia, "Automatic scene text recognition using a convolutional neural network", In Workshop on Camera-Based Document Analysis and Recognition, 2007.
 Yaoqun Xu, "Effect of white noise on chaotic neural network", Control and Decision Conference, CCDC'09, pp.3229-3234, 2009.
 R. M. Zur, Yulei Jiang, L. L. Pesce, K. Drukker, "Noise injection for training artificial neural networks: A comparation with weight decay and early stopping", Medical Physics, vol.36(10), pp.4810-4818, 2009.
[CrossRef] [Web of Science Times Cited 19] [SCOPUS Times Cited 24]
 F. Mamedov, Jamal Fathi Abu Hasna, "Character Recognition using Neural Networks", The 2006 World Congress in Computer Science, Computer Engineering and Applied Computing, ICAI06, 2006.
 G. Montavon, G. B. Orr, K. R. Muller, Neural Networks Tricks of the Trade, Springer-Verlag, LNCS7700, ISBN:978-3-642-35288-1, 2012.
 Yingqiao Shi, Wenbing Fan, Guodong Shi, "The research of printed character recognition based on neural network", Fourth International Symposium on Parallel Architecture, Algoritms and Programming, pp.119-122, 2011.
[CrossRef] [SCOPUS Times Cited 3]
 Li Fuliang, Gao Shuangxi, "Character recognition system board on backpropagation neural network", International Conference on Machine Vision and Human-machine Interface, pp.393-396, 2010.
 S. Geman, E. Bienenstock, R. Doursat, "Neural networks and the bias/variance dilemma", Neural Computation 4, pp.1-58, 1992.
[CrossRef] [Web of Science Times Cited 1230]
 M. I. Jordan, C. M. Bishop, "Neural Networks", ACM Computing Surveys, ISSN:0360-0300, 1996.
[CrossRef] [Web of Science Times Cited 13]
 A. Coates, H. Lee, A. Y. Ng, "An analysis of single layer networks in unsupervised feature learning", In AIS-TATS, 2011.
 A. I. Galushkin, Neural networks theory, ISBN: 978-3-540-48124-9, Springer-Verlag Berlin Heidelberg, 2007.
 S. Gheorghita, R. Munteanu, M. Enache, "Study of Neural Networks to Improve Performance for Character Recognition", Automation Quality and Testing Robotics (AQTR), IEEE International Conference, p. 323-326, 2012.
[CrossRef] [SCOPUS Times Cited 3]
 M. Hogan, H. Demuth, M. Beale, Neural network toolbox 6 users guide, 2008.
Web of Science® Citations for all references: 1,344 TCR
SCOPUS® Citations for all references: 121 TCR
Web of Science® Average Citations per reference: 84 ACR
SCOPUS® Average Citations per reference: 8 ACR
TCR = Total Citations for References / ACR = Average Citations per Reference
We introduced in 2010 - for the first time in scientific publishing, the term "References Weight", as a quantitative indication of the quality ... Read more
Citations for references updated on 2017-09-18 10:07 in 47 seconds.
Note1: Web of Science® is a registered trademark of Thomson Reuters.
Note2: SCOPUS® is a registered trademark of Elsevier B.V.
Disclaimer: All queries to the respective databases were made by using the DOI record of every reference (where available). Due to technical problems beyond our control, the information is not always accurate. Please use the CrossRef link to visit the respective publisher site.
Faculty of Electrical Engineering and Computer Science
Stefan cel Mare University of Suceava, Romania
All rights reserved: Advances in Electrical and Computer Engineering is a registered trademark of the Stefan cel Mare University of Suceava. No part of this publication may be reproduced, stored in a retrieval system, photocopied, recorded or archived, without the written permission from the Editor. When authors submit their papers for publication, they agree that the copyright for their article be transferred to the Faculty of Electrical Engineering and Computer Science, Stefan cel Mare University of Suceava, Romania, if and only if the articles are accepted for publication. The copyright covers the exclusive rights to reproduce and distribute the article, including reprints and translations.
Permission for other use: The copyright owner's consent does not extend to copying for general distribution, for promotion, for creating new works, or for resale. Specific written permission must be obtained from the Editor for such copying. Direct linking to files hosted on this website is strictly prohibited.
Disclaimer: Whilst every effort is made by the publishers and editorial board to see that no inaccurate or misleading data, opinions or statements appear in this journal, they wish to make it clear that all information and opinions formulated in the articles, as well as linguistic accuracy, are the sole responsibility of the author.