Comparative study of part-based handwritten character recognition methods

Wang Song, Seiichi Uchida, Marcus Liwicki

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

14 Citations (Scopus)

Abstract

The purpose of this paper is to introduce three part-based methods for handwritten character recognition and then compare their performances experimentally. All of those methods decompose handwritten characters into "parts". Then some recognition processes are done in a part-wise manner and, finally, the recognition results at all the parts are combined via voting to have the recognition result of the entire character. Since part-based methods do not rely on the global structure of the character, we can expect their robustness against various deformations. Three voting methods have been investigated for the combination: single voting, multiple voting, and class distance. All of them use different strategies for voting. Experimental results on the MNIST database showed the relative superiority of the class distance method and the robustness of the multiple voting method against the reduction of training set.

Original languageEnglish
Title of host publicationProceedings - 11th International Conference on Document Analysis and Recognition, ICDAR 2011
Pages814-818
Number of pages5
DOIs
Publication statusPublished - Dec 2 2011
Event11th International Conference on Document Analysis and Recognition, ICDAR 2011 - Beijing, China
Duration: Sep 18 2011Sep 21 2011

Publication series

NameProceedings of the International Conference on Document Analysis and Recognition, ICDAR
ISSN (Print)1520-5363

Other

Other11th International Conference on Document Analysis and Recognition, ICDAR 2011
CountryChina
CityBeijing
Period9/18/119/21/11

Fingerprint

Character recognition

All Science Journal Classification (ASJC) codes

  • Computer Vision and Pattern Recognition

Cite this

Song, W., Uchida, S., & Liwicki, M. (2011). Comparative study of part-based handwritten character recognition methods. In Proceedings - 11th International Conference on Document Analysis and Recognition, ICDAR 2011 (pp. 814-818). [6065424] (Proceedings of the International Conference on Document Analysis and Recognition, ICDAR). https://doi.org/10.1109/ICDAR.2011.167

Comparative study of part-based handwritten character recognition methods. / Song, Wang; Uchida, Seiichi; Liwicki, Marcus.

Proceedings - 11th International Conference on Document Analysis and Recognition, ICDAR 2011. 2011. p. 814-818 6065424 (Proceedings of the International Conference on Document Analysis and Recognition, ICDAR).

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

Song, W, Uchida, S & Liwicki, M 2011, Comparative study of part-based handwritten character recognition methods. in Proceedings - 11th International Conference on Document Analysis and Recognition, ICDAR 2011., 6065424, Proceedings of the International Conference on Document Analysis and Recognition, ICDAR, pp. 814-818, 11th International Conference on Document Analysis and Recognition, ICDAR 2011, Beijing, China, 9/18/11. https://doi.org/10.1109/ICDAR.2011.167
Song W, Uchida S, Liwicki M. Comparative study of part-based handwritten character recognition methods. In Proceedings - 11th International Conference on Document Analysis and Recognition, ICDAR 2011. 2011. p. 814-818. 6065424. (Proceedings of the International Conference on Document Analysis and Recognition, ICDAR). https://doi.org/10.1109/ICDAR.2011.167
Song, Wang ; Uchida, Seiichi ; Liwicki, Marcus. / Comparative study of part-based handwritten character recognition methods. Proceedings - 11th International Conference on Document Analysis and Recognition, ICDAR 2011. 2011. pp. 814-818 (Proceedings of the International Conference on Document Analysis and Recognition, ICDAR).
@inproceedings{0117491f0cb34640963cca7a34dec16d,
title = "Comparative study of part-based handwritten character recognition methods",
abstract = "The purpose of this paper is to introduce three part-based methods for handwritten character recognition and then compare their performances experimentally. All of those methods decompose handwritten characters into {"}parts{"}. Then some recognition processes are done in a part-wise manner and, finally, the recognition results at all the parts are combined via voting to have the recognition result of the entire character. Since part-based methods do not rely on the global structure of the character, we can expect their robustness against various deformations. Three voting methods have been investigated for the combination: single voting, multiple voting, and class distance. All of them use different strategies for voting. Experimental results on the MNIST database showed the relative superiority of the class distance method and the robustness of the multiple voting method against the reduction of training set.",
author = "Wang Song and Seiichi Uchida and Marcus Liwicki",
year = "2011",
month = "12",
day = "2",
doi = "10.1109/ICDAR.2011.167",
language = "English",
isbn = "9780769545202",
series = "Proceedings of the International Conference on Document Analysis and Recognition, ICDAR",
pages = "814--818",
booktitle = "Proceedings - 11th International Conference on Document Analysis and Recognition, ICDAR 2011",

}

TY - GEN

T1 - Comparative study of part-based handwritten character recognition methods

AU - Song, Wang

AU - Uchida, Seiichi

AU - Liwicki, Marcus

PY - 2011/12/2

Y1 - 2011/12/2

N2 - The purpose of this paper is to introduce three part-based methods for handwritten character recognition and then compare their performances experimentally. All of those methods decompose handwritten characters into "parts". Then some recognition processes are done in a part-wise manner and, finally, the recognition results at all the parts are combined via voting to have the recognition result of the entire character. Since part-based methods do not rely on the global structure of the character, we can expect their robustness against various deformations. Three voting methods have been investigated for the combination: single voting, multiple voting, and class distance. All of them use different strategies for voting. Experimental results on the MNIST database showed the relative superiority of the class distance method and the robustness of the multiple voting method against the reduction of training set.

AB - The purpose of this paper is to introduce three part-based methods for handwritten character recognition and then compare their performances experimentally. All of those methods decompose handwritten characters into "parts". Then some recognition processes are done in a part-wise manner and, finally, the recognition results at all the parts are combined via voting to have the recognition result of the entire character. Since part-based methods do not rely on the global structure of the character, we can expect their robustness against various deformations. Three voting methods have been investigated for the combination: single voting, multiple voting, and class distance. All of them use different strategies for voting. Experimental results on the MNIST database showed the relative superiority of the class distance method and the robustness of the multiple voting method against the reduction of training set.

UR - http://www.scopus.com/inward/record.url?scp=82355182360&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=82355182360&partnerID=8YFLogxK

U2 - 10.1109/ICDAR.2011.167

DO - 10.1109/ICDAR.2011.167

M3 - Conference contribution

AN - SCOPUS:82355182360

SN - 9780769545202

T3 - Proceedings of the International Conference on Document Analysis and Recognition, ICDAR

SP - 814

EP - 818

BT - Proceedings - 11th International Conference on Document Analysis and Recognition, ICDAR 2011

ER -