Skeleton extraction

Comparison of five methods on the Arabic IFN/ENIT database

Atallah M. Al-Shatnawi, Bader M. Alfawwaz, Khairuddin Omar, Ahmed M. Zeki

Research output: Chapter in Book/Report/Conference proceedingConference contribution

4 Citations (Scopus)

Abstract

Thinning 'Skeletonization' is a very crucial stage in the Arabic Character Recognition (ACR) system. It simplifies the text shape and reduces the amount of data that needs to be handled and it is usually used as a pre-processing stage for recognition and storage systems. The skeleton of Arabic text can be used for: baseline detection, character segmentation, and features extraction, and ultimately supporting the classification. In this paper, five of the state of the art thinning algorithms are selected and implemented. The five algorithms are: SPTA, Zhang-Suen parallel thinning algorithm, Huang-Wan-Liu thinning algorithm, thinning and skeletonization based morphological operation algorithms. The five selected algorithms are applied on the IFN/ENIT dataset. The results obtained by the five methods are discussed and analyzed against the IFN/ENIT dataset based on preserving shape and the text connectivity, preventing spurious tails, maintaining one pixel width skeleton and avoiding the necking problem as well as running time efficiently. In addition to that some performance measurement for checking text connectivity, spurious tails and calculating the stroke thickness are proposed and carried out.

Original languageEnglish
Title of host publication2014 6th International Conference on Computer Science and Information Technology, CSIT 2014 - Proceedings
PublisherIEEE Computer Society
Pages50-59
Number of pages10
DOIs
Publication statusPublished - 2014
Event2014 6th International Conference on Computer Science and Information Technology, CSIT 2014 - Amman
Duration: 26 Mar 201427 Mar 2014

Other

Other2014 6th International Conference on Computer Science and Information Technology, CSIT 2014
CityAmman
Period26/3/1427/3/14

Fingerprint

Character recognition
Feature extraction
Pixels
Processing

Keywords

  • Arabic Character Recognition
  • Huang-Wan-L
  • morphological
  • Skeleton
  • SPTA
  • Text connectivity
  • Thinning
  • Zhang-Suen

ASJC Scopus subject areas

  • Computational Theory and Mathematics
  • Information Systems

Cite this

Al-Shatnawi, A. M., Alfawwaz, B. M., Omar, K., & Zeki, A. M. (2014). Skeleton extraction: Comparison of five methods on the Arabic IFN/ENIT database. In 2014 6th International Conference on Computer Science and Information Technology, CSIT 2014 - Proceedings (pp. 50-59). [6805978] IEEE Computer Society. https://doi.org/10.1109/CSIT.2014.6805978

Skeleton extraction : Comparison of five methods on the Arabic IFN/ENIT database. / Al-Shatnawi, Atallah M.; Alfawwaz, Bader M.; Omar, Khairuddin; Zeki, Ahmed M.

2014 6th International Conference on Computer Science and Information Technology, CSIT 2014 - Proceedings. IEEE Computer Society, 2014. p. 50-59 6805978.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Al-Shatnawi, AM, Alfawwaz, BM, Omar, K & Zeki, AM 2014, Skeleton extraction: Comparison of five methods on the Arabic IFN/ENIT database. in 2014 6th International Conference on Computer Science and Information Technology, CSIT 2014 - Proceedings., 6805978, IEEE Computer Society, pp. 50-59, 2014 6th International Conference on Computer Science and Information Technology, CSIT 2014, Amman, 26/3/14. https://doi.org/10.1109/CSIT.2014.6805978
Al-Shatnawi AM, Alfawwaz BM, Omar K, Zeki AM. Skeleton extraction: Comparison of five methods on the Arabic IFN/ENIT database. In 2014 6th International Conference on Computer Science and Information Technology, CSIT 2014 - Proceedings. IEEE Computer Society. 2014. p. 50-59. 6805978 https://doi.org/10.1109/CSIT.2014.6805978
Al-Shatnawi, Atallah M. ; Alfawwaz, Bader M. ; Omar, Khairuddin ; Zeki, Ahmed M. / Skeleton extraction : Comparison of five methods on the Arabic IFN/ENIT database. 2014 6th International Conference on Computer Science and Information Technology, CSIT 2014 - Proceedings. IEEE Computer Society, 2014. pp. 50-59
@inproceedings{f543f8dc458f4e7cac919b0a4b57be0c,
title = "Skeleton extraction: Comparison of five methods on the Arabic IFN/ENIT database",
abstract = "Thinning 'Skeletonization' is a very crucial stage in the Arabic Character Recognition (ACR) system. It simplifies the text shape and reduces the amount of data that needs to be handled and it is usually used as a pre-processing stage for recognition and storage systems. The skeleton of Arabic text can be used for: baseline detection, character segmentation, and features extraction, and ultimately supporting the classification. In this paper, five of the state of the art thinning algorithms are selected and implemented. The five algorithms are: SPTA, Zhang-Suen parallel thinning algorithm, Huang-Wan-Liu thinning algorithm, thinning and skeletonization based morphological operation algorithms. The five selected algorithms are applied on the IFN/ENIT dataset. The results obtained by the five methods are discussed and analyzed against the IFN/ENIT dataset based on preserving shape and the text connectivity, preventing spurious tails, maintaining one pixel width skeleton and avoiding the necking problem as well as running time efficiently. In addition to that some performance measurement for checking text connectivity, spurious tails and calculating the stroke thickness are proposed and carried out.",
keywords = "Arabic Character Recognition, Huang-Wan-L, morphological, Skeleton, SPTA, Text connectivity, Thinning, Zhang-Suen",
author = "Al-Shatnawi, {Atallah M.} and Alfawwaz, {Bader M.} and Khairuddin Omar and Zeki, {Ahmed M.}",
year = "2014",
doi = "10.1109/CSIT.2014.6805978",
language = "English",
pages = "50--59",
booktitle = "2014 6th International Conference on Computer Science and Information Technology, CSIT 2014 - Proceedings",
publisher = "IEEE Computer Society",

}

TY - GEN

T1 - Skeleton extraction

T2 - Comparison of five methods on the Arabic IFN/ENIT database

AU - Al-Shatnawi, Atallah M.

AU - Alfawwaz, Bader M.

AU - Omar, Khairuddin

AU - Zeki, Ahmed M.

PY - 2014

Y1 - 2014

N2 - Thinning 'Skeletonization' is a very crucial stage in the Arabic Character Recognition (ACR) system. It simplifies the text shape and reduces the amount of data that needs to be handled and it is usually used as a pre-processing stage for recognition and storage systems. The skeleton of Arabic text can be used for: baseline detection, character segmentation, and features extraction, and ultimately supporting the classification. In this paper, five of the state of the art thinning algorithms are selected and implemented. The five algorithms are: SPTA, Zhang-Suen parallel thinning algorithm, Huang-Wan-Liu thinning algorithm, thinning and skeletonization based morphological operation algorithms. The five selected algorithms are applied on the IFN/ENIT dataset. The results obtained by the five methods are discussed and analyzed against the IFN/ENIT dataset based on preserving shape and the text connectivity, preventing spurious tails, maintaining one pixel width skeleton and avoiding the necking problem as well as running time efficiently. In addition to that some performance measurement for checking text connectivity, spurious tails and calculating the stroke thickness are proposed and carried out.

AB - Thinning 'Skeletonization' is a very crucial stage in the Arabic Character Recognition (ACR) system. It simplifies the text shape and reduces the amount of data that needs to be handled and it is usually used as a pre-processing stage for recognition and storage systems. The skeleton of Arabic text can be used for: baseline detection, character segmentation, and features extraction, and ultimately supporting the classification. In this paper, five of the state of the art thinning algorithms are selected and implemented. The five algorithms are: SPTA, Zhang-Suen parallel thinning algorithm, Huang-Wan-Liu thinning algorithm, thinning and skeletonization based morphological operation algorithms. The five selected algorithms are applied on the IFN/ENIT dataset. The results obtained by the five methods are discussed and analyzed against the IFN/ENIT dataset based on preserving shape and the text connectivity, preventing spurious tails, maintaining one pixel width skeleton and avoiding the necking problem as well as running time efficiently. In addition to that some performance measurement for checking text connectivity, spurious tails and calculating the stroke thickness are proposed and carried out.

KW - Arabic Character Recognition

KW - Huang-Wan-L

KW - morphological

KW - Skeleton

KW - SPTA

KW - Text connectivity

KW - Thinning

KW - Zhang-Suen

UR - http://www.scopus.com/inward/record.url?scp=84901200565&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84901200565&partnerID=8YFLogxK

U2 - 10.1109/CSIT.2014.6805978

DO - 10.1109/CSIT.2014.6805978

M3 - Conference contribution

SP - 50

EP - 59

BT - 2014 6th International Conference on Computer Science and Information Technology, CSIT 2014 - Proceedings

PB - IEEE Computer Society

ER -