Recognition of Japanese historical hand-written characters based on object detection methods

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

Abstract

We consider the recognition problem of Japanese historical handwritten characters called “Kuzushiji”. Unlike modern characters, Kuzushiji characters are harder to recognize partly because many of them are connected and not separated by spaces without any segmentation information. We propose two methods for segmentation and recognition of Kuzushiji characters. The first method learns segmentation rules and character classifiers simultaneously from data sets with character labels and segmentation information. Second method is for segmentation and can be used with any single character recognizer. Our methods outperform other baselines and achieve the state-of-the-art accuracy on both segmentation and recognition tasks on data sets of three consecutive Kuzushiji characters.

Original languageEnglish
Title of host publicationProceedings of the 2019 Workshop on Historical Document Imaging and Processing, HIP 2019
PublisherAssociation for Computing Machinery
Pages72-77
Number of pages6
ISBN (Electronic)9781450376686
DOIs
Publication statusPublished - Sep 20 2019
Event5th International Workshop on Historical Document Imaging and Processing, HIP 2019, held in conjunction with ICDAR 2019 - Sydney, Australia
Duration: Sep 20 2019Sep 21 2019

Publication series

NameACM International Conference Proceeding Series

Conference

Conference5th International Workshop on Historical Document Imaging and Processing, HIP 2019, held in conjunction with ICDAR 2019
CountryAustralia
CitySydney
Period9/20/199/21/19

Fingerprint

Labels
Classifiers
Object detection

All Science Journal Classification (ASJC) codes

  • Human-Computer Interaction
  • Computer Networks and Communications
  • Computer Vision and Pattern Recognition
  • Software

Cite this

Tang, Y., Hatano, K., & Takimoto, E. (2019). Recognition of Japanese historical hand-written characters based on object detection methods. In Proceedings of the 2019 Workshop on Historical Document Imaging and Processing, HIP 2019 (pp. 72-77). (ACM International Conference Proceeding Series). Association for Computing Machinery. https://doi.org/10.1145/3352631.3352642

Recognition of Japanese historical hand-written characters based on object detection methods. / Tang, Yiping; Hatano, Kohei; Takimoto, Eiji.

Proceedings of the 2019 Workshop on Historical Document Imaging and Processing, HIP 2019. Association for Computing Machinery, 2019. p. 72-77 (ACM International Conference Proceeding Series).

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

Tang, Y, Hatano, K & Takimoto, E 2019, Recognition of Japanese historical hand-written characters based on object detection methods. in Proceedings of the 2019 Workshop on Historical Document Imaging and Processing, HIP 2019. ACM International Conference Proceeding Series, Association for Computing Machinery, pp. 72-77, 5th International Workshop on Historical Document Imaging and Processing, HIP 2019, held in conjunction with ICDAR 2019, Sydney, Australia, 9/20/19. https://doi.org/10.1145/3352631.3352642
Tang Y, Hatano K, Takimoto E. Recognition of Japanese historical hand-written characters based on object detection methods. In Proceedings of the 2019 Workshop on Historical Document Imaging and Processing, HIP 2019. Association for Computing Machinery. 2019. p. 72-77. (ACM International Conference Proceeding Series). https://doi.org/10.1145/3352631.3352642
Tang, Yiping ; Hatano, Kohei ; Takimoto, Eiji. / Recognition of Japanese historical hand-written characters based on object detection methods. Proceedings of the 2019 Workshop on Historical Document Imaging and Processing, HIP 2019. Association for Computing Machinery, 2019. pp. 72-77 (ACM International Conference Proceeding Series).
@inproceedings{f836e6e98c294d77bb0df1ba1bc8f91a,
title = "Recognition of Japanese historical hand-written characters based on object detection methods",
abstract = "We consider the recognition problem of Japanese historical handwritten characters called “Kuzushiji”. Unlike modern characters, Kuzushiji characters are harder to recognize partly because many of them are connected and not separated by spaces without any segmentation information. We propose two methods for segmentation and recognition of Kuzushiji characters. The first method learns segmentation rules and character classifiers simultaneously from data sets with character labels and segmentation information. Second method is for segmentation and can be used with any single character recognizer. Our methods outperform other baselines and achieve the state-of-the-art accuracy on both segmentation and recognition tasks on data sets of three consecutive Kuzushiji characters.",
author = "Yiping Tang and Kohei Hatano and Eiji Takimoto",
year = "2019",
month = "9",
day = "20",
doi = "10.1145/3352631.3352642",
language = "English",
series = "ACM International Conference Proceeding Series",
publisher = "Association for Computing Machinery",
pages = "72--77",
booktitle = "Proceedings of the 2019 Workshop on Historical Document Imaging and Processing, HIP 2019",

}

TY - GEN

T1 - Recognition of Japanese historical hand-written characters based on object detection methods

AU - Tang, Yiping

AU - Hatano, Kohei

AU - Takimoto, Eiji

PY - 2019/9/20

Y1 - 2019/9/20

N2 - We consider the recognition problem of Japanese historical handwritten characters called “Kuzushiji”. Unlike modern characters, Kuzushiji characters are harder to recognize partly because many of them are connected and not separated by spaces without any segmentation information. We propose two methods for segmentation and recognition of Kuzushiji characters. The first method learns segmentation rules and character classifiers simultaneously from data sets with character labels and segmentation information. Second method is for segmentation and can be used with any single character recognizer. Our methods outperform other baselines and achieve the state-of-the-art accuracy on both segmentation and recognition tasks on data sets of three consecutive Kuzushiji characters.

AB - We consider the recognition problem of Japanese historical handwritten characters called “Kuzushiji”. Unlike modern characters, Kuzushiji characters are harder to recognize partly because many of them are connected and not separated by spaces without any segmentation information. We propose two methods for segmentation and recognition of Kuzushiji characters. The first method learns segmentation rules and character classifiers simultaneously from data sets with character labels and segmentation information. Second method is for segmentation and can be used with any single character recognizer. Our methods outperform other baselines and achieve the state-of-the-art accuracy on both segmentation and recognition tasks on data sets of three consecutive Kuzushiji characters.

UR - http://www.scopus.com/inward/record.url?scp=85074780655&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85074780655&partnerID=8YFLogxK

U2 - 10.1145/3352631.3352642

DO - 10.1145/3352631.3352642

M3 - Conference contribution

AN - SCOPUS:85074780655

T3 - ACM International Conference Proceeding Series

SP - 72

EP - 77

BT - Proceedings of the 2019 Workshop on Historical Document Imaging and Processing, HIP 2019

PB - Association for Computing Machinery

ER -