Enhancing the Ensemble-Based Scene Character Recognition by Using Classification Likelihood

Fuma Horie; Hideaki Goto; Takuo Suganuma

doi:10.1007/978-3-030-41299-9_10

Enhancing the Ensemble-Based Scene Character Recognition by Using Classification Likelihood

Fuma Horie, Hideaki Goto, Takuo Suganuma

Research and Development Divisions

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

Abstract

Research on scene character recognition has been popular for its potential in many applications including automatic translator, signboard recognition, and reading assistant for the visually-impaired. The scene character recognition is challenging and difficult owing to various environmental factors at image capturing and complex design of characters. Current OCR systems have not gained practical accuracy for arbitrary scene characters, although some effective methods were proposed in the past. In order to enhance existing recognition systems, we propose a hierarchical recognition method utilizing the classification likelihood and image pre-processing methods. It is shown that the accuracy of our latest ensemble system has been improved from 80.7% to 82.3% by adopting the proposed methods.

Original language	English
Title of host publication	Pattern Recognition - 5th Asian Conference, ACPR 2019, Revised Selected Papers
Editors	Shivakumara Palaiahnakote, Gabriella Sanniti di Baja, Liang Wang, Wei Qi Yan
Publisher	Springer
Pages	118-128
Number of pages	11
ISBN (Print)	9783030412982
DOIs	https://doi.org/10.1007/978-3-030-41299-9_10
Publication status	Published - 2020
Event	5th Asian Conference on Pattern Recognition, ACPR 2019 - Auckland, New Zealand Duration: 2019 Nov 26 → 2019 Nov 29

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	12047 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	5th Asian Conference on Pattern Recognition, ACPR 2019
Country/Territory	New Zealand
City	Auckland
Period	19/11/26 → 19/11/29

Keywords

Ensemble voting classifier
Hierarchical recognition method
Synthetic Scene Character Data

ASJC Scopus subject areas

Theoretical Computer Science
Computer Science(all)

Access to Document

10.1007/978-3-030-41299-9_10

Cite this

Horie, F., Goto, H., & Suganuma, T. (2020). Enhancing the Ensemble-Based Scene Character Recognition by Using Classification Likelihood. In S. Palaiahnakote, G. Sanniti di Baja, L. Wang, & W. Q. Yan (Eds.), Pattern Recognition - 5th Asian Conference, ACPR 2019, Revised Selected Papers (pp. 118-128). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 12047 LNCS). Springer. https://doi.org/10.1007/978-3-030-41299-9_10

Enhancing the Ensemble-Based Scene Character Recognition by Using Classification Likelihood. / Horie, Fuma; Goto, Hideaki ; Suganuma, Takuo.
Pattern Recognition - 5th Asian Conference, ACPR 2019, Revised Selected Papers. ed. / Shivakumara Palaiahnakote; Gabriella Sanniti di Baja; Liang Wang; Wei Qi Yan. Springer, 2020. p. 118-128 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 12047 LNCS).

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

Horie, F, Goto, H & Suganuma, T 2020, Enhancing the Ensemble-Based Scene Character Recognition by Using Classification Likelihood. in S Palaiahnakote, G Sanniti di Baja, L Wang & WQ Yan (eds), Pattern Recognition - 5th Asian Conference, ACPR 2019, Revised Selected Papers. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 12047 LNCS, Springer, pp. 118-128, 5th Asian Conference on Pattern Recognition, ACPR 2019, Auckland, New Zealand, 19/11/26. https://doi.org/10.1007/978-3-030-41299-9_10

Horie F, Goto H , Suganuma T. Enhancing the Ensemble-Based Scene Character Recognition by Using Classification Likelihood. In Palaiahnakote S, Sanniti di Baja G, Wang L, Yan WQ, editors, Pattern Recognition - 5th Asian Conference, ACPR 2019, Revised Selected Papers. Springer. 2020. p. 118-128. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-41299-9_10

Horie, Fuma ; Goto, Hideaki ; Suganuma, Takuo. / Enhancing the Ensemble-Based Scene Character Recognition by Using Classification Likelihood. Pattern Recognition - 5th Asian Conference, ACPR 2019, Revised Selected Papers. editor / Shivakumara Palaiahnakote ; Gabriella Sanniti di Baja ; Liang Wang ; Wei Qi Yan. Springer, 2020. pp. 118-128 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{7a7f8b8bbf0d470da06c761a73981771,

title = "Enhancing the Ensemble-Based Scene Character Recognition by Using Classification Likelihood",

abstract = "Research on scene character recognition has been popular for its potential in many applications including automatic translator, signboard recognition, and reading assistant for the visually-impaired. The scene character recognition is challenging and difficult owing to various environmental factors at image capturing and complex design of characters. Current OCR systems have not gained practical accuracy for arbitrary scene characters, although some effective methods were proposed in the past. In order to enhance existing recognition systems, we propose a hierarchical recognition method utilizing the classification likelihood and image pre-processing methods. It is shown that the accuracy of our latest ensemble system has been improved from 80.7% to 82.3% by adopting the proposed methods.",

keywords = "Ensemble voting classifier, Hierarchical recognition method, Synthetic Scene Character Data",

author = "Fuma Horie and Hideaki Goto and Takuo Suganuma",

note = "Publisher Copyright: {\textcopyright} 2020, Springer Nature Switzerland AG.; 5th Asian Conference on Pattern Recognition, ACPR 2019 ; Conference date: 26-11-2019 Through 29-11-2019",

year = "2020",

doi = "10.1007/978-3-030-41299-9_10",

language = "English",

isbn = "9783030412982",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer",

pages = "118--128",

editor = "Shivakumara Palaiahnakote and {Sanniti di Baja}, Gabriella and Liang Wang and Yan, {Wei Qi}",

booktitle = "Pattern Recognition - 5th Asian Conference, ACPR 2019, Revised Selected Papers",

}

TY - GEN

T1 - Enhancing the Ensemble-Based Scene Character Recognition by Using Classification Likelihood

AU - Horie, Fuma

AU - Goto, Hideaki

AU - Suganuma, Takuo

PY - 2020

Y1 - 2020

N2 - Research on scene character recognition has been popular for its potential in many applications including automatic translator, signboard recognition, and reading assistant for the visually-impaired. The scene character recognition is challenging and difficult owing to various environmental factors at image capturing and complex design of characters. Current OCR systems have not gained practical accuracy for arbitrary scene characters, although some effective methods were proposed in the past. In order to enhance existing recognition systems, we propose a hierarchical recognition method utilizing the classification likelihood and image pre-processing methods. It is shown that the accuracy of our latest ensemble system has been improved from 80.7% to 82.3% by adopting the proposed methods.

AB - Research on scene character recognition has been popular for its potential in many applications including automatic translator, signboard recognition, and reading assistant for the visually-impaired. The scene character recognition is challenging and difficult owing to various environmental factors at image capturing and complex design of characters. Current OCR systems have not gained practical accuracy for arbitrary scene characters, although some effective methods were proposed in the past. In order to enhance existing recognition systems, we propose a hierarchical recognition method utilizing the classification likelihood and image pre-processing methods. It is shown that the accuracy of our latest ensemble system has been improved from 80.7% to 82.3% by adopting the proposed methods.

KW - Ensemble voting classifier

KW - Hierarchical recognition method

KW - Synthetic Scene Character Data

UR - http://www.scopus.com/inward/record.url?scp=85081578373&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85081578373&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-41299-9_10

DO - 10.1007/978-3-030-41299-9_10

M3 - Conference contribution

AN - SCOPUS:85081578373

SN - 9783030412982

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

SP - 118

EP - 128

BT - Pattern Recognition - 5th Asian Conference, ACPR 2019, Revised Selected Papers

A2 - Palaiahnakote, Shivakumara

A2 - Sanniti di Baja, Gabriella

A2 - Wang, Liang

A2 - Yan, Wei Qi

PB - Springer

T2 - 5th Asian Conference on Pattern Recognition, ACPR 2019

Y2 - 26 November 2019 through 29 November 2019

ER -

Enhancing the Ensemble-Based Scene Character Recognition by Using Classification Likelihood

Abstract

Publication series

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this