A three-level classifier: Fuzzy C Means, Support Vector Machine and unique pixels for Arabic handwritten digits

Maen Takruri; Rami Al-Hmouz; Ahmed Al-Hmouz

doi:10.1109/WSCAR.2014.6916798

A three-level classifier: Fuzzy C Means, Support Vector Machine and unique pixels for Arabic handwritten digits

Maen Takruri^*, Rami Al-Hmouz, Ahmed Al-Hmouz

^*Corresponding author for this work

Electrical and Computer Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

7 Citations (Scopus)

Abstract

In this study, we present a classification approach for handwritten Arabic digits (symbols). Like numbers in other languages, Arabic numbers consists of nine digits. Character images of Arabic digits are similar in the sense that one single classifier will not give a reliable classification rate. Therefore, the implementation of more levels of classification is important for the realization. We introduce Fuzzy C-Means based classifier for the lower level and Support Vector Machine SVM for the second level when more details are required and finally confirmation of classification will be through unique pixels. The unique pixel method forms the third classification level. It works on determining the pixel areas that are unique to each digit. The unique pixel method decision is compared with decision of the lower classifier (FCM) and top classifier (SVM). The algorithm is tested on 3510 images. The overall testing accuracy reported is 88%.

Original language	English
Title of host publication	2014 World Symposium on Computer Applications and Research, WSCAR 2014
Publisher	Institute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)	9781479928057
DOIs	https://doi.org/10.1109/WSCAR.2014.6916798
Publication status	Published - Oct 3 2014
Event	2014 World Symposium on Computer Applications and Research, WSCAR 2014 - Sousse, Tunisia Duration: Jan 18 2014 → Jan 20 2014

Conference

Conference	2014 World Symposium on Computer Applications and Research, WSCAR 2014
Country/Territory	Tunisia
City	Sousse
Period	1/18/14 → 1/20/14

Keywords

Arabic numbers
Fuzzy- C Mean
OCR
SVM

ASJC Scopus subject areas

Computer Science Applications

Access to Document

10.1109/WSCAR.2014.6916798

Cite this

A three-level classifier: Fuzzy C Means, Support Vector Machine and unique pixels for Arabic handwritten digits. / Takruri, Maen; Al-Hmouz, Rami; Al-Hmouz, Ahmed.
2014 World Symposium on Computer Applications and Research, WSCAR 2014. Institute of Electrical and Electronics Engineers Inc., 2014. 6916798.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Takruri, M, Al-Hmouz, R & Al-Hmouz, A 2014, A three-level classifier: Fuzzy C Means, Support Vector Machine and unique pixels for Arabic handwritten digits. in 2014 World Symposium on Computer Applications and Research, WSCAR 2014., 6916798, Institute of Electrical and Electronics Engineers Inc., 2014 World Symposium on Computer Applications and Research, WSCAR 2014, Sousse, Tunisia, 1/18/14. https://doi.org/10.1109/WSCAR.2014.6916798

@inproceedings{21922f5d6aaa4653974b68de1a00bd13,

title = "A three-level classifier: Fuzzy C Means, Support Vector Machine and unique pixels for Arabic handwritten digits",

abstract = "In this study, we present a classification approach for handwritten Arabic digits (symbols). Like numbers in other languages, Arabic numbers consists of nine digits. Character images of Arabic digits are similar in the sense that one single classifier will not give a reliable classification rate. Therefore, the implementation of more levels of classification is important for the realization. We introduce Fuzzy C-Means based classifier for the lower level and Support Vector Machine SVM for the second level when more details are required and finally confirmation of classification will be through unique pixels. The unique pixel method forms the third classification level. It works on determining the pixel areas that are unique to each digit. The unique pixel method decision is compared with decision of the lower classifier (FCM) and top classifier (SVM). The algorithm is tested on 3510 images. The overall testing accuracy reported is 88%.",

keywords = "Arabic numbers, Fuzzy- C Mean, OCR, SVM",

author = "Maen Takruri and Rami Al-Hmouz and Ahmed Al-Hmouz",

note = "Publisher Copyright: {\textcopyright} 2014 IEEE.; 2014 World Symposium on Computer Applications and Research, WSCAR 2014 ; Conference date: 18-01-2014 Through 20-01-2014",

year = "2014",

month = oct,

day = "3",

doi = "10.1109/WSCAR.2014.6916798",

language = "English",

booktitle = "2014 World Symposium on Computer Applications and Research, WSCAR 2014",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

address = "United States",

}

TY - GEN

T1 - A three-level classifier

T2 - 2014 World Symposium on Computer Applications and Research, WSCAR 2014

AU - Takruri, Maen

AU - Al-Hmouz, Rami

AU - Al-Hmouz, Ahmed

PY - 2014/10/3

Y1 - 2014/10/3

N2 - In this study, we present a classification approach for handwritten Arabic digits (symbols). Like numbers in other languages, Arabic numbers consists of nine digits. Character images of Arabic digits are similar in the sense that one single classifier will not give a reliable classification rate. Therefore, the implementation of more levels of classification is important for the realization. We introduce Fuzzy C-Means based classifier for the lower level and Support Vector Machine SVM for the second level when more details are required and finally confirmation of classification will be through unique pixels. The unique pixel method forms the third classification level. It works on determining the pixel areas that are unique to each digit. The unique pixel method decision is compared with decision of the lower classifier (FCM) and top classifier (SVM). The algorithm is tested on 3510 images. The overall testing accuracy reported is 88%.

AB - In this study, we present a classification approach for handwritten Arabic digits (symbols). Like numbers in other languages, Arabic numbers consists of nine digits. Character images of Arabic digits are similar in the sense that one single classifier will not give a reliable classification rate. Therefore, the implementation of more levels of classification is important for the realization. We introduce Fuzzy C-Means based classifier for the lower level and Support Vector Machine SVM for the second level when more details are required and finally confirmation of classification will be through unique pixels. The unique pixel method forms the third classification level. It works on determining the pixel areas that are unique to each digit. The unique pixel method decision is compared with decision of the lower classifier (FCM) and top classifier (SVM). The algorithm is tested on 3510 images. The overall testing accuracy reported is 88%.

KW - Arabic numbers

KW - Fuzzy- C Mean

KW - OCR

KW - SVM

UR - http://www.scopus.com/inward/record.url?scp=84908884628&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84908884628&partnerID=8YFLogxK

U2 - 10.1109/WSCAR.2014.6916798

DO - 10.1109/WSCAR.2014.6916798

M3 - Conference contribution

AN - SCOPUS:84908884628

BT - 2014 World Symposium on Computer Applications and Research, WSCAR 2014

PB - Institute of Electrical and Electronics Engineers Inc.

Y2 - 18 January 2014 through 20 January 2014

ER -

A three-level classifier: Fuzzy C Means, Support Vector Machine and unique pixels for Arabic handwritten digits

Abstract

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this