Scene text recognition algorithm based on faster RCNN

Boya Wang; Jianqing Xu; Junbao Li; Cong Hu; Jeng Shyang Pan

doi:10.1109/EIIS.2017.8298720

Scene text recognition algorithm based on faster RCNN

Boya Wang
, Jianqing Xu
, Junbao Li
, Cong Hu
, Jeng Shyang Pan

Faculty of Computing

Harbin Institute of Technology
Guilin University of Electronic Technology
Fujian University of Technology

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

10 Scopus citations

Abstract

Industrial session of the natural scene in the text recognition technology has a great demand. The traditional optical character recognition technology (OCR) requires the text neat layout and neatness and background clean, and industrial production often fail to meet such standards. In this paper, a new text recognition algorithm based on deep learning is proposed for the existing problems of OCR technology. In this paper, a new method based on convolution neural network (Faster RCNN) is proposed to improve the correctness of text recognition. Compared with the conventional detection method, the correct rate of recognition based on Faster RCNN model can reach 90.4%, and the correctness rate is 88.9%. Experiments show that the recognition method in this paper is effective.

Original language	English
Title of host publication	1st International Conference on Electronics Instrumentation and Information Systems, EIIS 2017
Editors	Jun-Bao Li
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	1-4
Number of pages	4
ISBN (Electronic)	9781538608432
DOIs	https://doi.org/10.1109/EIIS.2017.8298720
State	Published - 2 Jul 2017
Event	1st International Conference on Electronics Instrumentation and Information Systems, EIIS 2017 - Harbin, China Duration: 3 Jun 2017 → 5 Jun 2017

Publication series

Name	1st International Conference on Electronics Instrumentation and Information Systems, EIIS 2017
Volume	2018-January

Conference

Conference	1st International Conference on Electronics Instrumentation and Information Systems, EIIS 2017
Country/Territory	China
City	Harbin
Period	3/06/17 → 5/06/17

Keywords

convolution neural network
deep learning
scene text recognition

Access to Document

10.1109/EIIS.2017.8298720

Cite this

Wang, B., Xu, J., Li, J., Hu, C., & Pan, J. S. (2017). Scene text recognition algorithm based on faster RCNN. In J.-B. Li (Ed.), 1st International Conference on Electronics Instrumentation and Information Systems, EIIS 2017 (pp. 1-4). (1st International Conference on Electronics Instrumentation and Information Systems, EIIS 2017; Vol. 2018-January). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/EIIS.2017.8298720

Wang, Boya ; Xu, Jianqing ; Li, Junbao et al. / Scene text recognition algorithm based on faster RCNN. 1st International Conference on Electronics Instrumentation and Information Systems, EIIS 2017. editor / Jun-Bao Li. Institute of Electrical and Electronics Engineers Inc., 2017. pp. 1-4 (1st International Conference on Electronics Instrumentation and Information Systems, EIIS 2017).

@inproceedings{9809a017f71049658d26e023929c9660,

title = "Scene text recognition algorithm based on faster RCNN",

abstract = "Industrial session of the natural scene in the text recognition technology has a great demand. The traditional optical character recognition technology (OCR) requires the text neat layout and neatness and background clean, and industrial production often fail to meet such standards. In this paper, a new text recognition algorithm based on deep learning is proposed for the existing problems of OCR technology. In this paper, a new method based on convolution neural network (Faster RCNN) is proposed to improve the correctness of text recognition. Compared with the conventional detection method, the correct rate of recognition based on Faster RCNN model can reach 90.4\%, and the correctness rate is 88.9\%. Experiments show that the recognition method in this paper is effective.",

keywords = "convolution neural network, deep learning, scene text recognition",

author = "Boya Wang and Jianqing Xu and Junbao Li and Cong Hu and Pan, \{Jeng Shyang\}",

note = "Publisher Copyright: {\textcopyright} 2017 IEEE.; 1st International Conference on Electronics Instrumentation and Information Systems, EIIS 2017 ; Conference date: 03-06-2017 Through 05-06-2017",

year = "2017",

month = jul,

day = "2",

doi = "10.1109/EIIS.2017.8298720",

language = "英语",

series = "1st International Conference on Electronics Instrumentation and Information Systems, EIIS 2017",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "1--4",

editor = "Jun-Bao Li",

booktitle = "1st International Conference on Electronics Instrumentation and Information Systems, EIIS 2017",

address = "美国",

}

Wang, B, Xu, J, Li, J, Hu, C & Pan, JS 2017, Scene text recognition algorithm based on faster RCNN. in J-B Li (ed.), 1st International Conference on Electronics Instrumentation and Information Systems, EIIS 2017. 1st International Conference on Electronics Instrumentation and Information Systems, EIIS 2017, vol. 2018-January, Institute of Electrical and Electronics Engineers Inc., pp. 1-4, 1st International Conference on Electronics Instrumentation and Information Systems, EIIS 2017, Harbin, China, 3/06/17. https://doi.org/10.1109/EIIS.2017.8298720

Scene text recognition algorithm based on faster RCNN. / Wang, Boya; Xu, Jianqing; Li, Junbao et al.
1st International Conference on Electronics Instrumentation and Information Systems, EIIS 2017. ed. / Jun-Bao Li. Institute of Electrical and Electronics Engineers Inc., 2017. p. 1-4 (1st International Conference on Electronics Instrumentation and Information Systems, EIIS 2017; Vol. 2018-January).

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

TY - GEN

T1 - Scene text recognition algorithm based on faster RCNN

AU - Wang, Boya

AU - Xu, Jianqing

AU - Li, Junbao

AU - Hu, Cong

AU - Pan, Jeng Shyang

PY - 2017/7/2

Y1 - 2017/7/2

N2 - Industrial session of the natural scene in the text recognition technology has a great demand. The traditional optical character recognition technology (OCR) requires the text neat layout and neatness and background clean, and industrial production often fail to meet such standards. In this paper, a new text recognition algorithm based on deep learning is proposed for the existing problems of OCR technology. In this paper, a new method based on convolution neural network (Faster RCNN) is proposed to improve the correctness of text recognition. Compared with the conventional detection method, the correct rate of recognition based on Faster RCNN model can reach 90.4%, and the correctness rate is 88.9%. Experiments show that the recognition method in this paper is effective.

AB - Industrial session of the natural scene in the text recognition technology has a great demand. The traditional optical character recognition technology (OCR) requires the text neat layout and neatness and background clean, and industrial production often fail to meet such standards. In this paper, a new text recognition algorithm based on deep learning is proposed for the existing problems of OCR technology. In this paper, a new method based on convolution neural network (Faster RCNN) is proposed to improve the correctness of text recognition. Compared with the conventional detection method, the correct rate of recognition based on Faster RCNN model can reach 90.4%, and the correctness rate is 88.9%. Experiments show that the recognition method in this paper is effective.

KW - convolution neural network

KW - deep learning

KW - scene text recognition

UR - https://www.scopus.com/pages/publications/85048810899

U2 - 10.1109/EIIS.2017.8298720

DO - 10.1109/EIIS.2017.8298720

M3 - 会议稿件

AN - SCOPUS:85048810899

T3 - 1st International Conference on Electronics Instrumentation and Information Systems, EIIS 2017

SP - 1

EP - 4

BT - 1st International Conference on Electronics Instrumentation and Information Systems, EIIS 2017

A2 - Li, Jun-Bao

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 1st International Conference on Electronics Instrumentation and Information Systems, EIIS 2017

Y2 - 3 June 2017 through 5 June 2017

ER -

Wang B, Xu J, Li J, Hu C, Pan JS. Scene text recognition algorithm based on faster RCNN. In Li JB, editor, 1st International Conference on Electronics Instrumentation and Information Systems, EIIS 2017. Institute of Electrical and Electronics Engineers Inc. 2017. p. 1-4. (1st International Conference on Electronics Instrumentation and Information Systems, EIIS 2017). doi: 10.1109/EIIS.2017.8298720

Scene text recognition algorithm based on faster RCNN

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this