Smooth stroke width transform for text detection

Il Seok Oh; Jin Seon Lee

doi:10.1007/978-3-319-44748-3_18

Smooth stroke width transform for text detection

Il Seok Oh^*
, Jin Seon Lee

^*Corresponding author for this work

Woosuk University

Research output: Contribution to conference › Conference paper › peer-review

5 Scopus citations

Abstract

The stroke width transform (SWT) is a generic operation for the task of detecting texts from natural images because the characters intrinsically have the elongated shape of nearly uniform width. The edge pairing technique was recently developed by Epshtein et al. and is popularly used due to its simplicity and effectiveness. However since the natural images are noisy and sensitive to variations, high degree of artifacts arises and it hinders subsequent processing of the text detection. This paper reformulates the SWT problem in a new way that searches for an optimal solution in 3-D space. We present an effective search algorithm called the aggregation approach, borrowed from the depth image reconstruction domain. The experiments showed that the algorithm produced a smooth SWT map which is better for subsequent processes.

Original language	English
Title of host publication	Artificial Intelligence
Subtitle of host publication	Methodology, Systems, and Applications - 17th International Conference, AIMSA 2016, Proceedings
Editors	Gennady Agre, Christo Dichev
Publisher	Springer Verlag
Pages	183-191
Number of pages	9
ISBN (Print)	9783319447476
DOIs	https://doi.org/10.1007/978-3-319-44748-3_18
State	Published - 2016
Event	17th International Conference on Artificial Intelligence: Methodology, Systems, and Applications, AIMSA 2016 - Varna, Bulgaria Duration: 2016.09.7 → 2016.09.10

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	9883 LNAI
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	17th International Conference on Artificial Intelligence: Methodology, Systems, and Applications, AIMSA 2016
Country/Territory	Bulgaria
City	Varna
Period	16.09.7 → 16.09.10

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 3 Good Health and Well-being

Keywords

Natural scene recognition
Regularization
Stroke width transform
Text detection

Quacquarelli Symonds(QS) Subject Topics

Computer Science & Information Systems

Access to Document

10.1007/978-3-319-44748-3_18

Cite this

Oh, I. S., & Lee, J. S. (2016). Smooth stroke width transform for text detection. In G. Agre, & C. Dichev (Eds.), Artificial Intelligence: Methodology, Systems, and Applications - 17th International Conference, AIMSA 2016, Proceedings (pp. 183-191). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9883 LNAI). Springer Verlag. https://doi.org/10.1007/978-3-319-44748-3_18

Oh, Il Seok ; Lee, Jin Seon. / Smooth stroke width transform for text detection. Artificial Intelligence: Methodology, Systems, and Applications - 17th International Conference, AIMSA 2016, Proceedings. editor / Gennady Agre ; Christo Dichev. Springer Verlag, 2016. pp. 183-191 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{747c5a3bb875425f96851bde708194eb,

title = "Smooth stroke width transform for text detection",

abstract = "The stroke width transform (SWT) is a generic operation for the task of detecting texts from natural images because the characters intrinsically have the elongated shape of nearly uniform width. The edge pairing technique was recently developed by Epshtein et al. and is popularly used due to its simplicity and effectiveness. However since the natural images are noisy and sensitive to variations, high degree of artifacts arises and it hinders subsequent processing of the text detection. This paper reformulates the SWT problem in a new way that searches for an optimal solution in 3-D space. We present an effective search algorithm called the aggregation approach, borrowed from the depth image reconstruction domain. The experiments showed that the algorithm produced a smooth SWT map which is better for subsequent processes.",

keywords = "Natural scene recognition, Regularization, Stroke width transform, Text detection",

author = "Oh, \{Il Seok\} and Lee, \{Jin Seon\}",

note = "Publisher Copyright: {\textcopyright} Springer International Publishing Switzerland 2016.; 17th International Conference on Artificial Intelligence: Methodology, Systems, and Applications, AIMSA 2016 ; Conference date: 07-09-2016 Through 10-09-2016",

year = "2016",

doi = "10.1007/978-3-319-44748-3\_18",

language = "English",

isbn = "9783319447476",

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

publisher = "Springer Verlag",

pages = "183--191",

editor = "Gennady Agre and Christo Dichev",

booktitle = "Artificial Intelligence",

}

Oh, IS & Lee, JS 2016, Smooth stroke width transform for text detection. in G Agre & C Dichev (eds), Artificial Intelligence: Methodology, Systems, and Applications - 17th International Conference, AIMSA 2016, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 9883 LNAI, Springer Verlag, pp. 183-191, 17th International Conference on Artificial Intelligence: Methodology, Systems, and Applications, AIMSA 2016, Varna, Bulgaria, 16.09.7. https://doi.org/10.1007/978-3-319-44748-3_18

Smooth stroke width transform for text detection. / Oh, Il Seok; Lee, Jin Seon.
Artificial Intelligence: Methodology, Systems, and Applications - 17th International Conference, AIMSA 2016, Proceedings. ed. / Gennady Agre; Christo Dichev. Springer Verlag, 2016. p. 183-191 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9883 LNAI).

Research output: Contribution to conference › Conference paper › peer-review

TY - GEN

T1 - Smooth stroke width transform for text detection

AU - Oh, Il Seok

AU - Lee, Jin Seon

N1 - Publisher Copyright: © Springer International Publishing Switzerland 2016.

PY - 2016

Y1 - 2016

N2 - The stroke width transform (SWT) is a generic operation for the task of detecting texts from natural images because the characters intrinsically have the elongated shape of nearly uniform width. The edge pairing technique was recently developed by Epshtein et al. and is popularly used due to its simplicity and effectiveness. However since the natural images are noisy and sensitive to variations, high degree of artifacts arises and it hinders subsequent processing of the text detection. This paper reformulates the SWT problem in a new way that searches for an optimal solution in 3-D space. We present an effective search algorithm called the aggregation approach, borrowed from the depth image reconstruction domain. The experiments showed that the algorithm produced a smooth SWT map which is better for subsequent processes.

AB - The stroke width transform (SWT) is a generic operation for the task of detecting texts from natural images because the characters intrinsically have the elongated shape of nearly uniform width. The edge pairing technique was recently developed by Epshtein et al. and is popularly used due to its simplicity and effectiveness. However since the natural images are noisy and sensitive to variations, high degree of artifacts arises and it hinders subsequent processing of the text detection. This paper reformulates the SWT problem in a new way that searches for an optimal solution in 3-D space. We present an effective search algorithm called the aggregation approach, borrowed from the depth image reconstruction domain. The experiments showed that the algorithm produced a smooth SWT map which is better for subsequent processes.

KW - Natural scene recognition

KW - Regularization

KW - Stroke width transform

KW - Text detection

UR - https://www.scopus.com/pages/publications/84986206043

U2 - 10.1007/978-3-319-44748-3_18

DO - 10.1007/978-3-319-44748-3_18

M3 - Conference paper

AN - SCOPUS:84986206043

SN - 9783319447476

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

SP - 183

EP - 191

BT - Artificial Intelligence

A2 - Agre, Gennady

A2 - Dichev, Christo

PB - Springer Verlag

T2 - 17th International Conference on Artificial Intelligence: Methodology, Systems, and Applications, AIMSA 2016

Y2 - 7 September 2016 through 10 September 2016

ER -

Oh IS, Lee JS. Smooth stroke width transform for text detection. In Agre G, Dichev C, editors, Artificial Intelligence: Methodology, Systems, and Applications - 17th International Conference, AIMSA 2016, Proceedings. Springer Verlag. 2016. p. 183-191. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-44748-3_18

Smooth stroke width transform for text detection

Abstract

Publication series

Conference

UN SDGs

Keywords

Quacquarelli Symonds(QS) Subject Topics

Access to Document

Other files and links

Fingerprint

Cite this