Improved YOLOv3 Infrared Image Pedestrian Detection Algorithm

Jianting Shi; Guiqiang Zhang; Jie Yuan; Yingtao Zhang

doi:10.1007/978-981-15-7981-3_37

Improved YOLOv3 Infrared Image Pedestrian Detection Algorithm

Jianting Shi^*
, Guiqiang Zhang
, Jie Yuan
, Yingtao Zhang

^*Corresponding author for this work

Heilongjiang University of Science and Technology
Shanghai Aerospace Electronic Technology Institute
School of Computer Science and Technology, Harbin Institute of Technology

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

6 Scopus citations

Abstract

Security surveillance is widely used in daily life. For nighttime or complicated monitoring environments, this article proposes an infrared pedestrian monitoring based on YOLOv3. In the original YOLOv3 network structure, two aspects of optimization were made. One was to optimize the scale in the residual structure, and the rich features of the deconvolution layer were added to the original residual structure. The other was to use the DenseNet network to enhance the features. The optimization of fusion ability and delivery ability effectively improves the detection ability for small targets, and the pedestrian detection performance based on infrared images. After comparative testing, compared with YOLOv3, the overall mean average precision is improved by 4.39% to 78.86%.

Original language	English
Title of host publication	Data Science - 6th International Conference of Pioneering Computer Scientists, Engineers and Educators, ICPCSEE 2020, Proceedings
Editors	Jianchao Zeng, Weipeng Jing, Xianhua Song, Zeguang Lu
Publisher	Springer
Pages	506-517
Number of pages	12
ISBN (Print)	9789811579806
DOIs	https://doi.org/10.1007/978-981-15-7981-3_37
State	Published - 2020
Externally published	Yes
Event	6th International Conference of Pioneering Computer Scientists, Engineers and Educators, ICPCSEE 2020 - Taiyuan, China Duration: 18 Sep 2020 → 21 Sep 2020

Publication series

Name	Communications in Computer and Information Science
Volume	1257 CCIS
ISSN (Print)	1865-0929
ISSN (Electronic)	1865-0937

Conference

Conference	6th International Conference of Pioneering Computer Scientists, Engineers and Educators, ICPCSEE 2020
Country/Territory	China
City	Taiyuan
Period	18/09/20 → 21/09/20

Keywords

DenseNet
Infrared
YOLOv3

Access to Document

10.1007/978-981-15-7981-3_37

Cite this

Shi, J., Zhang, G., Yuan, J., & Zhang, Y. (2020). Improved YOLOv3 Infrared Image Pedestrian Detection Algorithm. In J. Zeng, W. Jing, X. Song, & Z. Lu (Eds.), Data Science - 6th International Conference of Pioneering Computer Scientists, Engineers and Educators, ICPCSEE 2020, Proceedings (pp. 506-517). (Communications in Computer and Information Science; Vol. 1257 CCIS). Springer. https://doi.org/10.1007/978-981-15-7981-3_37

Shi, Jianting ; Zhang, Guiqiang ; Yuan, Jie et al. / Improved YOLOv3 Infrared Image Pedestrian Detection Algorithm. Data Science - 6th International Conference of Pioneering Computer Scientists, Engineers and Educators, ICPCSEE 2020, Proceedings. editor / Jianchao Zeng ; Weipeng Jing ; Xianhua Song ; Zeguang Lu. Springer, 2020. pp. 506-517 (Communications in Computer and Information Science).

@inproceedings{3c0354c503fe41139f86e21879a5fd73,

title = "Improved YOLOv3 Infrared Image Pedestrian Detection Algorithm",

abstract = "Security surveillance is widely used in daily life. For nighttime or complicated monitoring environments, this article proposes an infrared pedestrian monitoring based on YOLOv3. In the original YOLOv3 network structure, two aspects of optimization were made. One was to optimize the scale in the residual structure, and the rich features of the deconvolution layer were added to the original residual structure. The other was to use the DenseNet network to enhance the features. The optimization of fusion ability and delivery ability effectively improves the detection ability for small targets, and the pedestrian detection performance based on infrared images. After comparative testing, compared with YOLOv3, the overall mean average precision is improved by 4.39\% to 78.86\%.",

keywords = "DenseNet, Infrared, YOLOv3",

author = "Jianting Shi and Guiqiang Zhang and Jie Yuan and Yingtao Zhang",

note = "Publisher Copyright: {\textcopyright} 2020, Springer Nature Singapore Pte Ltd.; 6th International Conference of Pioneering Computer Scientists, Engineers and Educators, ICPCSEE 2020 ; Conference date: 18-09-2020 Through 21-09-2020",

year = "2020",

doi = "10.1007/978-981-15-7981-3\_37",

language = "英语",

isbn = "9789811579806",

series = "Communications in Computer and Information Science",

publisher = "Springer",

pages = "506--517",

editor = "Jianchao Zeng and Weipeng Jing and Xianhua Song and Zeguang Lu",

booktitle = "Data Science - 6th International Conference of Pioneering Computer Scientists, Engineers and Educators, ICPCSEE 2020, Proceedings",

address = "德国",

}

Shi, J, Zhang, G, Yuan, J & Zhang, Y 2020, Improved YOLOv3 Infrared Image Pedestrian Detection Algorithm. in J Zeng, W Jing, X Song & Z Lu (eds), Data Science - 6th International Conference of Pioneering Computer Scientists, Engineers and Educators, ICPCSEE 2020, Proceedings. Communications in Computer and Information Science, vol. 1257 CCIS, Springer, pp. 506-517, 6th International Conference of Pioneering Computer Scientists, Engineers and Educators, ICPCSEE 2020, Taiyuan, China, 18/09/20. https://doi.org/10.1007/978-981-15-7981-3_37

Improved YOLOv3 Infrared Image Pedestrian Detection Algorithm. / Shi, Jianting; Zhang, Guiqiang; Yuan, Jie et al.
Data Science - 6th International Conference of Pioneering Computer Scientists, Engineers and Educators, ICPCSEE 2020, Proceedings. ed. / Jianchao Zeng; Weipeng Jing; Xianhua Song; Zeguang Lu. Springer, 2020. p. 506-517 (Communications in Computer and Information Science; Vol. 1257 CCIS).

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

TY - GEN

T1 - Improved YOLOv3 Infrared Image Pedestrian Detection Algorithm

AU - Shi, Jianting

AU - Zhang, Guiqiang

AU - Yuan, Jie

AU - Zhang, Yingtao

PY - 2020

Y1 - 2020

N2 - Security surveillance is widely used in daily life. For nighttime or complicated monitoring environments, this article proposes an infrared pedestrian monitoring based on YOLOv3. In the original YOLOv3 network structure, two aspects of optimization were made. One was to optimize the scale in the residual structure, and the rich features of the deconvolution layer were added to the original residual structure. The other was to use the DenseNet network to enhance the features. The optimization of fusion ability and delivery ability effectively improves the detection ability for small targets, and the pedestrian detection performance based on infrared images. After comparative testing, compared with YOLOv3, the overall mean average precision is improved by 4.39% to 78.86%.

AB - Security surveillance is widely used in daily life. For nighttime or complicated monitoring environments, this article proposes an infrared pedestrian monitoring based on YOLOv3. In the original YOLOv3 network structure, two aspects of optimization were made. One was to optimize the scale in the residual structure, and the rich features of the deconvolution layer were added to the original residual structure. The other was to use the DenseNet network to enhance the features. The optimization of fusion ability and delivery ability effectively improves the detection ability for small targets, and the pedestrian detection performance based on infrared images. After comparative testing, compared with YOLOv3, the overall mean average precision is improved by 4.39% to 78.86%.

KW - DenseNet

KW - Infrared

KW - YOLOv3

UR - https://www.scopus.com/pages/publications/85090036660

U2 - 10.1007/978-981-15-7981-3_37

DO - 10.1007/978-981-15-7981-3_37

M3 - 会议稿件

AN - SCOPUS:85090036660

SN - 9789811579806

T3 - Communications in Computer and Information Science

SP - 506

EP - 517

BT - Data Science - 6th International Conference of Pioneering Computer Scientists, Engineers and Educators, ICPCSEE 2020, Proceedings

A2 - Zeng, Jianchao

A2 - Jing, Weipeng

A2 - Song, Xianhua

A2 - Lu, Zeguang

PB - Springer

T2 - 6th International Conference of Pioneering Computer Scientists, Engineers and Educators, ICPCSEE 2020

Y2 - 18 September 2020 through 21 September 2020

ER -

Shi J, Zhang G, Yuan J, Zhang Y. Improved YOLOv3 Infrared Image Pedestrian Detection Algorithm. In Zeng J, Jing W, Song X, Lu Z, editors, Data Science - 6th International Conference of Pioneering Computer Scientists, Engineers and Educators, ICPCSEE 2020, Proceedings. Springer. 2020. p. 506-517. (Communications in Computer and Information Science). doi: 10.1007/978-981-15-7981-3_37

Improved YOLOv3 Infrared Image Pedestrian Detection Algorithm

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this