Simultaneous mixed vertical and horizontal handwritten Japanese character line detection

Tomotaka Kimura, Chinthaka Premachandra, Hiroharu Kawanaka

研究成果: Conference contribution

抄録

Teachers consume considerable time and their energy in process of marking examination sheets. To reduce this burden on teachers, we have been developing an automatic marking system. To mark examination sheets automatically, handwritten character lines are extracted, and then the characters on those lines are recognized. In this paper, we discuss how character lines are extracted from Japanese handwritten examination sheets without ruled lines. Japanese characters can be written vertically and horizontally, so examination sheets written in Japanese are consisted of mixed vertical and horizontal (MVH) character lines. Conventional character line extraction methods cannot deal with MVH lines, because they have been developed to consider only horizontal character lines. This paper focuses on the simultaneous detection of MVH character lines. The result of experiments using appropriate examination sheet images shows that our method can detect MVH character lines effectively.

元の言語English
ホスト出版物のタイトルComputer Vision and Graphics - International Conference, ICCVG 2016, Proceedings
出版者Springer Verlag
ページ564-572
ページ数9
9972 LNCS
ISBN(印刷物)9783319464176
DOI
出版物ステータスPublished - 2016
イベントInternational Conference on Computer Vision and Graphics, ICCVG 2016 - Warsaw, Poland
継続期間: 2016 9 192016 9 21

出版物シリーズ

名前Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
9972 LNCS
ISSN(印刷物)03029743
ISSN(電子版)16113349

Other

OtherInternational Conference on Computer Vision and Graphics, ICCVG 2016
Poland
Warsaw
期間16/9/1916/9/21

Fingerprint

Experiments

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

これを引用

Kimura, T., Premachandra, C., & Kawanaka, H. (2016). Simultaneous mixed vertical and horizontal handwritten Japanese character line detection. : Computer Vision and Graphics - International Conference, ICCVG 2016, Proceedings (巻 9972 LNCS, pp. 564-572). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); 巻数 9972 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-46418-3_50

Simultaneous mixed vertical and horizontal handwritten Japanese character line detection. / Kimura, Tomotaka; Premachandra, Chinthaka; Kawanaka, Hiroharu.

Computer Vision and Graphics - International Conference, ICCVG 2016, Proceedings. 巻 9972 LNCS Springer Verlag, 2016. p. 564-572 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); 巻 9972 LNCS).

研究成果: Conference contribution

Kimura, T, Premachandra, C & Kawanaka, H 2016, Simultaneous mixed vertical and horizontal handwritten Japanese character line detection. : Computer Vision and Graphics - International Conference, ICCVG 2016, Proceedings. 巻. 9972 LNCS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 巻. 9972 LNCS, Springer Verlag, pp. 564-572, International Conference on Computer Vision and Graphics, ICCVG 2016, Warsaw, Poland, 16/9/19. https://doi.org/10.1007/978-3-319-46418-3_50
Kimura T, Premachandra C, Kawanaka H. Simultaneous mixed vertical and horizontal handwritten Japanese character line detection. : Computer Vision and Graphics - International Conference, ICCVG 2016, Proceedings. 巻 9972 LNCS. Springer Verlag. 2016. p. 564-572. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-46418-3_50
Kimura, Tomotaka ; Premachandra, Chinthaka ; Kawanaka, Hiroharu. / Simultaneous mixed vertical and horizontal handwritten Japanese character line detection. Computer Vision and Graphics - International Conference, ICCVG 2016, Proceedings. 巻 9972 LNCS Springer Verlag, 2016. pp. 564-572 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{75b028d0ba37408cb2147417a3902549,
title = "Simultaneous mixed vertical and horizontal handwritten Japanese character line detection",
abstract = "Teachers consume considerable time and their energy in process of marking examination sheets. To reduce this burden on teachers, we have been developing an automatic marking system. To mark examination sheets automatically, handwritten character lines are extracted, and then the characters on those lines are recognized. In this paper, we discuss how character lines are extracted from Japanese handwritten examination sheets without ruled lines. Japanese characters can be written vertically and horizontally, so examination sheets written in Japanese are consisted of mixed vertical and horizontal (MVH) character lines. Conventional character line extraction methods cannot deal with MVH lines, because they have been developed to consider only horizontal character lines. This paper focuses on the simultaneous detection of MVH character lines. The result of experiments using appropriate examination sheet images shows that our method can detect MVH character lines effectively.",
author = "Tomotaka Kimura and Chinthaka Premachandra and Hiroharu Kawanaka",
year = "2016",
doi = "10.1007/978-3-319-46418-3_50",
language = "English",
isbn = "9783319464176",
volume = "9972 LNCS",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "564--572",
booktitle = "Computer Vision and Graphics - International Conference, ICCVG 2016, Proceedings",

}

TY - GEN

T1 - Simultaneous mixed vertical and horizontal handwritten Japanese character line detection

AU - Kimura, Tomotaka

AU - Premachandra, Chinthaka

AU - Kawanaka, Hiroharu

PY - 2016

Y1 - 2016

N2 - Teachers consume considerable time and their energy in process of marking examination sheets. To reduce this burden on teachers, we have been developing an automatic marking system. To mark examination sheets automatically, handwritten character lines are extracted, and then the characters on those lines are recognized. In this paper, we discuss how character lines are extracted from Japanese handwritten examination sheets without ruled lines. Japanese characters can be written vertically and horizontally, so examination sheets written in Japanese are consisted of mixed vertical and horizontal (MVH) character lines. Conventional character line extraction methods cannot deal with MVH lines, because they have been developed to consider only horizontal character lines. This paper focuses on the simultaneous detection of MVH character lines. The result of experiments using appropriate examination sheet images shows that our method can detect MVH character lines effectively.

AB - Teachers consume considerable time and their energy in process of marking examination sheets. To reduce this burden on teachers, we have been developing an automatic marking system. To mark examination sheets automatically, handwritten character lines are extracted, and then the characters on those lines are recognized. In this paper, we discuss how character lines are extracted from Japanese handwritten examination sheets without ruled lines. Japanese characters can be written vertically and horizontally, so examination sheets written in Japanese are consisted of mixed vertical and horizontal (MVH) character lines. Conventional character line extraction methods cannot deal with MVH lines, because they have been developed to consider only horizontal character lines. This paper focuses on the simultaneous detection of MVH character lines. The result of experiments using appropriate examination sheet images shows that our method can detect MVH character lines effectively.

UR - http://www.scopus.com/inward/record.url?scp=84989884549&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84989884549&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-46418-3_50

DO - 10.1007/978-3-319-46418-3_50

M3 - Conference contribution

AN - SCOPUS:84989884549

SN - 9783319464176

VL - 9972 LNCS

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 564

EP - 572

BT - Computer Vision and Graphics - International Conference, ICCVG 2016, Proceedings

PB - Springer Verlag

ER -