Localization of slab identification numbers using deep learning

Sang Jun Lee; Jaepil Ban; Hyeyeon Choi; Sang Woo Kim

doi:10.1109/ICCAS.2016.7832461

Localization of slab identification numbers using deep learning

Sang Jun Lee
, Jaepil Ban
, Hyeyeon Choi
, Sang Woo Kim

Division of Electronic Engineering

Pohang University of Science and Technology

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

5 Scopus citations

Abstract

In the steel industries, recognizing product information is an important task for the management of the manufacturing processes. For real factory scenes, localization of product identification numbers is conducted prior to recognition to obtain a satisfactory performance. The objective of this paper is localization of slab identification numbers in real factory scenes. Traditionally, most researches in the field of image processing and pattern recognition were focused on feature representation or shallow learning. However, conventional rule-based algorithms heavily depend on carefully engineered feature values and require heuristic parameter tuning. To overcome these limitations, a deep learning based algorithm is proposed for the localization with the minimum of manual interventions. This paper contains construction of training data, labeling process, and an architecture of a deep convolutional neural network. The performance error is remarkably reduced to 2.19% by the proposed algorithm compared to 4.59% in the previous work. By using a data-based method, this algorithm is easily expandable to apply for other applications.

Original language	English
Title of host publication	ICCAS 2016 - 2016 16th International Conference on Control, Automation and Systems, Proceedings
Publisher	IEEE Computer Society
Pages	1174-1176
Number of pages	3
ISBN (Electronic)	9788993215120
DOIs	https://doi.org/10.1109/ICCAS.2016.7832461
State	Published - 2016.01.24
Event	16th International Conference on Control, Automation and Systems, ICCAS 2016 - Gyeongju, Korea, Republic of Duration: 2016.10.16 → 2016.10.19

Publication series

Name	International Conference on Control, Automation and Systems
Volume	0
ISSN (Print)	1598-7833

Conference

Conference	16th International Conference on Control, Automation and Systems, ICCAS 2016
Country/Territory	Korea, Republic of
City	Gyeongju
Period	16.10.16 → 16.10.19

UN SDGs

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

SDG 9 Industry, Innovation, and Infrastructure

Keywords

deep convo-lutional neural network
deep learning
Industrial application
product identification number
steel slab
text localization

Access to Document

10.1109/ICCAS.2016.7832461

Cite this

Lee, S. J., Ban, J., Choi, H., & Kim, S. W. (2016). Localization of slab identification numbers using deep learning. In ICCAS 2016 - 2016 16th International Conference on Control, Automation and Systems, Proceedings (pp. 1174-1176). Article 7832461 (International Conference on Control, Automation and Systems; Vol. 0). IEEE Computer Society. https://doi.org/10.1109/ICCAS.2016.7832461

@inproceedings{c1a369e0f88542ed86536c2b3e34686c,

title = "Localization of slab identification numbers using deep learning",

abstract = "In the steel industries, recognizing product information is an important task for the management of the manufacturing processes. For real factory scenes, localization of product identification numbers is conducted prior to recognition to obtain a satisfactory performance. The objective of this paper is localization of slab identification numbers in real factory scenes. Traditionally, most researches in the field of image processing and pattern recognition were focused on feature representation or shallow learning. However, conventional rule-based algorithms heavily depend on carefully engineered feature values and require heuristic parameter tuning. To overcome these limitations, a deep learning based algorithm is proposed for the localization with the minimum of manual interventions. This paper contains construction of training data, labeling process, and an architecture of a deep convolutional neural network. The performance error is remarkably reduced to 2.19\% by the proposed algorithm compared to 4.59\% in the previous work. By using a data-based method, this algorithm is easily expandable to apply for other applications.",

keywords = "deep convo-lutional neural network, deep learning, Industrial application, product identification number, steel slab, text localization",

author = "Lee, \{Sang Jun\} and Jaepil Ban and Hyeyeon Choi and Kim, \{Sang Woo\}",

note = "Publisher Copyright: {\textcopyright} 2016 Institute of Control, Robotics and Systems - ICROS.; 16th International Conference on Control, Automation and Systems, ICCAS 2016 ; Conference date: 16-10-2016 Through 19-10-2016",

year = "2016",

month = jan,

day = "24",

doi = "10.1109/ICCAS.2016.7832461",

language = "English",

series = "International Conference on Control, Automation and Systems",

publisher = "IEEE Computer Society",

pages = "1174--1176",

booktitle = "ICCAS 2016 - 2016 16th International Conference on Control, Automation and Systems, Proceedings",

}

Lee, SJ, Ban, J, Choi, H & Kim, SW 2016, Localization of slab identification numbers using deep learning. in ICCAS 2016 - 2016 16th International Conference on Control, Automation and Systems, Proceedings., 7832461, International Conference on Control, Automation and Systems, vol. 0, IEEE Computer Society, pp. 1174-1176, 16th International Conference on Control, Automation and Systems, ICCAS 2016, Gyeongju, Korea, Republic of, 16.10.16. https://doi.org/10.1109/ICCAS.2016.7832461

Localization of slab identification numbers using deep learning. / Lee, Sang Jun; Ban, Jaepil; Choi, Hyeyeon et al.
ICCAS 2016 - 2016 16th International Conference on Control, Automation and Systems, Proceedings. IEEE Computer Society, 2016. p. 1174-1176 7832461 (International Conference on Control, Automation and Systems; Vol. 0).

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

TY - GEN

T1 - Localization of slab identification numbers using deep learning

AU - Lee, Sang Jun

AU - Ban, Jaepil

AU - Choi, Hyeyeon

AU - Kim, Sang Woo

PY - 2016/1/24

Y1 - 2016/1/24

N2 - In the steel industries, recognizing product information is an important task for the management of the manufacturing processes. For real factory scenes, localization of product identification numbers is conducted prior to recognition to obtain a satisfactory performance. The objective of this paper is localization of slab identification numbers in real factory scenes. Traditionally, most researches in the field of image processing and pattern recognition were focused on feature representation or shallow learning. However, conventional rule-based algorithms heavily depend on carefully engineered feature values and require heuristic parameter tuning. To overcome these limitations, a deep learning based algorithm is proposed for the localization with the minimum of manual interventions. This paper contains construction of training data, labeling process, and an architecture of a deep convolutional neural network. The performance error is remarkably reduced to 2.19% by the proposed algorithm compared to 4.59% in the previous work. By using a data-based method, this algorithm is easily expandable to apply for other applications.

AB - In the steel industries, recognizing product information is an important task for the management of the manufacturing processes. For real factory scenes, localization of product identification numbers is conducted prior to recognition to obtain a satisfactory performance. The objective of this paper is localization of slab identification numbers in real factory scenes. Traditionally, most researches in the field of image processing and pattern recognition were focused on feature representation or shallow learning. However, conventional rule-based algorithms heavily depend on carefully engineered feature values and require heuristic parameter tuning. To overcome these limitations, a deep learning based algorithm is proposed for the localization with the minimum of manual interventions. This paper contains construction of training data, labeling process, and an architecture of a deep convolutional neural network. The performance error is remarkably reduced to 2.19% by the proposed algorithm compared to 4.59% in the previous work. By using a data-based method, this algorithm is easily expandable to apply for other applications.

KW - deep convo-lutional neural network

KW - deep learning

KW - Industrial application

KW - product identification number

KW - steel slab

KW - text localization

UR - https://www.scopus.com/pages/publications/85013998325

U2 - 10.1109/ICCAS.2016.7832461

DO - 10.1109/ICCAS.2016.7832461

M3 - Conference paper

AN - SCOPUS:85013998325

T3 - International Conference on Control, Automation and Systems

SP - 1174

EP - 1176

BT - ICCAS 2016 - 2016 16th International Conference on Control, Automation and Systems, Proceedings

PB - IEEE Computer Society

T2 - 16th International Conference on Control, Automation and Systems, ICCAS 2016

Y2 - 16 October 2016 through 19 October 2016

ER -

Localization of slab identification numbers using deep learning

Abstract

Publication series

Conference

UN SDGs

Keywords

Access to Document

Other files and links

Fingerprint

Cite this