Street Sign Recognition Algorithm Based on Deep Learning

Jiawei Wang; Cen Chen; Chongwen Wang

doi:10.1145/3383812.3383819

Street Sign Recognition Algorithm Based on Deep Learning

Jiawei Wang, Cen Chen, Chongwen Wang

School of Computer Science and Technology

Beijing Institute of Technology

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

2 Citations (Scopus)

Abstract

The complex background, uneven illumination and object occlusion have increased the difficulty of scene texts detection. In this paper, we improved the existing object detection algorithm SSD, and made it possible to detect text objects in traffic guidance sign. We used a deep neural network CRNN to identify the text. This network is a combination of Convolution Neural Network and Recurrent Neural Network. At the same time, we proposed a new idea to optimize the detection algorithm through the text recognition result, so that the whole network can be trained end-to-end. According to the experimental results, the detection network achieves 88% mAP on our dataset at 11.6FPS, which has a good recognition effect.

Original language	English
Title of host publication	ICIGP 2020 - Proceedings of the 2020 3rd International Conference on Image and Graphics Processing
Publisher	Association for Computing Machinery
Pages	31-35
Number of pages	5
ISBN (Electronic)	9781450377201
DOIs	https://doi.org/10.1145/3383812.3383819
Publication status	Published - 8 Feb 2020
Event	3rd International Conference on Image and Graphics Processing, ICIGP 2020 - Singapore, Singapore Duration: 8 Feb 2020 → 10 Feb 2020

Publication series

Name	ACM International Conference Proceeding Series

Conference

Conference	3rd International Conference on Image and Graphics Processing, ICIGP 2020
Country/Territory	Singapore
City	Singapore
Period	8/02/20 → 10/02/20

Keywords

Deep Neural Network
Detection
Identification
Scene Text
Traffic Sign

Access to Document

10.1145/3383812.3383819

Cite this

@inproceedings{5fd8c83abe9a4788aaf2183759f4a759,

title = "Street Sign Recognition Algorithm Based on Deep Learning",

abstract = "The complex background, uneven illumination and object occlusion have increased the difficulty of scene texts detection. In this paper, we improved the existing object detection algorithm SSD, and made it possible to detect text objects in traffic guidance sign. We used a deep neural network CRNN to identify the text. This network is a combination of Convolution Neural Network and Recurrent Neural Network. At the same time, we proposed a new idea to optimize the detection algorithm through the text recognition result, so that the whole network can be trained end-to-end. According to the experimental results, the detection network achieves 88\% mAP on our dataset at 11.6FPS, which has a good recognition effect.",

keywords = "Deep Neural Network, Detection, Identification, Scene Text, Traffic Sign",

author = "Jiawei Wang and Cen Chen and Chongwen Wang",

note = "Publisher Copyright: {\textcopyright} 2020 ACM.; 3rd International Conference on Image and Graphics Processing, ICIGP 2020 ; Conference date: 08-02-2020 Through 10-02-2020",

year = "2020",

month = feb,

day = "8",

doi = "10.1145/3383812.3383819",

language = "English",

series = "ACM International Conference Proceeding Series",

publisher = "Association for Computing Machinery",

pages = "31--35",

booktitle = "ICIGP 2020 - Proceedings of the 2020 3rd International Conference on Image and Graphics Processing",

}

Wang, J, Chen, C & Wang, C 2020, Street Sign Recognition Algorithm Based on Deep Learning. in ICIGP 2020 - Proceedings of the 2020 3rd International Conference on Image and Graphics Processing. ACM International Conference Proceeding Series, Association for Computing Machinery, pp. 31-35, 3rd International Conference on Image and Graphics Processing, ICIGP 2020, Singapore, Singapore, 8/02/20. https://doi.org/10.1145/3383812.3383819

Street Sign Recognition Algorithm Based on Deep Learning. / Wang, Jiawei; Chen, Cen; Wang, Chongwen.
ICIGP 2020 - Proceedings of the 2020 3rd International Conference on Image and Graphics Processing. Association for Computing Machinery, 2020. p. 31-35 (ACM International Conference Proceeding Series).

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

TY - GEN

T1 - Street Sign Recognition Algorithm Based on Deep Learning

AU - Wang, Jiawei

AU - Chen, Cen

AU - Wang, Chongwen

PY - 2020/2/8

Y1 - 2020/2/8

N2 - The complex background, uneven illumination and object occlusion have increased the difficulty of scene texts detection. In this paper, we improved the existing object detection algorithm SSD, and made it possible to detect text objects in traffic guidance sign. We used a deep neural network CRNN to identify the text. This network is a combination of Convolution Neural Network and Recurrent Neural Network. At the same time, we proposed a new idea to optimize the detection algorithm through the text recognition result, so that the whole network can be trained end-to-end. According to the experimental results, the detection network achieves 88% mAP on our dataset at 11.6FPS, which has a good recognition effect.

AB - The complex background, uneven illumination and object occlusion have increased the difficulty of scene texts detection. In this paper, we improved the existing object detection algorithm SSD, and made it possible to detect text objects in traffic guidance sign. We used a deep neural network CRNN to identify the text. This network is a combination of Convolution Neural Network and Recurrent Neural Network. At the same time, we proposed a new idea to optimize the detection algorithm through the text recognition result, so that the whole network can be trained end-to-end. According to the experimental results, the detection network achieves 88% mAP on our dataset at 11.6FPS, which has a good recognition effect.

KW - Deep Neural Network

KW - Detection

KW - Identification

KW - Scene Text

KW - Traffic Sign

UR - http://www.scopus.com/inward/record.url?scp=85083107429&partnerID=8YFLogxK

U2 - 10.1145/3383812.3383819

DO - 10.1145/3383812.3383819

M3 - Conference contribution

AN - SCOPUS:85083107429

T3 - ACM International Conference Proceeding Series

SP - 31

EP - 35

BT - ICIGP 2020 - Proceedings of the 2020 3rd International Conference on Image and Graphics Processing

PB - Association for Computing Machinery

T2 - 3rd International Conference on Image and Graphics Processing, ICIGP 2020

Y2 - 8 February 2020 through 10 February 2020

ER -

Street Sign Recognition Algorithm Based on Deep Learning

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this