Research on Object Detection Methods in Low-Light Conditions

Feifan Wang; Xi’ai Chen; Xudong Wang; Weihong Ren; Yandong Tang

doi:10.1007/978-981-99-6492-5_48

Research on Object Detection Methods in Low-Light Conditions

Feifan Wang
, Xi’ai Chen^*
, Xudong Wang
, Weihong Ren
, Yandong Tang

^*Corresponding author for this work

Shenyang Ligong University
CAS - Shenyang Institute of Automation
Chinese Academy of Sciences
University of Chinese Academy of Sciences
Harbin Institute of Technology Shenzhen

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

5 Scopus citations

Abstract

Low-light images are images taken in poorly illuminated environments. Such images suffer from colour distortion, loss of detail and blurriness, which seriously affects the detection accuracy of object detection tasks. In order to improve the accuracy of object detection in low-light images, we propose a low-light image object detection algorithm based on image enhancement. The algorithm is jointly trained on the input side of the YOLOv5 network in combination with an unsupervised low-light enhancement model. The training phase optimises the overall network with the loss of object detection so that the image enhancement results are more favourable for improving the object detection accuracy. In the feature extraction phase, we design a feature enhancement model based on an attention mechanism. Our algorithm is tested on the publicly available ExDark dataset and achieves a mean average precision (mAP) of 79.15%, which is a 4.25% improvement over the baseline.

Original language	English
Title of host publication	Intelligent Robotics and Applications - 16th International Conference, ICIRA 2023, Proceedings
Editors	Huayong Yang, Honghai Liu, Jun Zou, Zhouping Yin, Lianqing Liu, Geng Yang, Xiaoping Ouyang, Zhiyong Wang
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	564-574
Number of pages	11
ISBN (Print)	9789819964918
DOIs	https://doi.org/10.1007/978-981-99-6492-5_48
State	Published - 2023
Externally published	Yes
Event	16th International Conference on Intelligent Robotics and Applications, ICIRA 2023 - Hangzhou, China Duration: 5 Jul 2023 → 7 Jul 2023

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	14270 LNAI
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	16th International Conference on Intelligent Robotics and Applications, ICIRA 2023
Country/Territory	China
City	Hangzhou
Period	5/07/23 → 7/07/23

Keywords

Attention mechanisms
Deep learning
Low-light enhancement
Object detection

Access to Document

10.1007/978-981-99-6492-5_48

Cite this

Wang, F., Chen, X., Wang, X., Ren, W., & Tang, Y. (2023). Research on Object Detection Methods in Low-Light Conditions. In H. Yang, H. Liu, J. Zou, Z. Yin, L. Liu, G. Yang, X. Ouyang, & Z. Wang (Eds.), Intelligent Robotics and Applications - 16th International Conference, ICIRA 2023, Proceedings (pp. 564-574). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 14270 LNAI). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-981-99-6492-5_48

Wang, Feifan ; Chen, Xi’ai ; Wang, Xudong et al. / Research on Object Detection Methods in Low-Light Conditions. Intelligent Robotics and Applications - 16th International Conference, ICIRA 2023, Proceedings. editor / Huayong Yang ; Honghai Liu ; Jun Zou ; Zhouping Yin ; Lianqing Liu ; Geng Yang ; Xiaoping Ouyang ; Zhiyong Wang. Springer Science and Business Media Deutschland GmbH, 2023. pp. 564-574 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{acf0fe76fb13477491e5709bd94c1cd2,

title = "Research on Object Detection Methods in Low-Light Conditions",

abstract = "Low-light images are images taken in poorly illuminated environments. Such images suffer from colour distortion, loss of detail and blurriness, which seriously affects the detection accuracy of object detection tasks. In order to improve the accuracy of object detection in low-light images, we propose a low-light image object detection algorithm based on image enhancement. The algorithm is jointly trained on the input side of the YOLOv5 network in combination with an unsupervised low-light enhancement model. The training phase optimises the overall network with the loss of object detection so that the image enhancement results are more favourable for improving the object detection accuracy. In the feature extraction phase, we design a feature enhancement model based on an attention mechanism. Our algorithm is tested on the publicly available ExDark dataset and achieves a mean average precision (mAP) of 79.15\%, which is a 4.25\% improvement over the baseline.",

keywords = "Attention mechanisms, Deep learning, Low-light enhancement, Object detection",

author = "Feifan Wang and Xi{\textquoteright}ai Chen and Xudong Wang and Weihong Ren and Yandong Tang",

note = "Publisher Copyright: {\textcopyright} The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd 2023.; 16th International Conference on Intelligent Robotics and Applications, ICIRA 2023 ; Conference date: 05-07-2023 Through 07-07-2023",

year = "2023",

doi = "10.1007/978-981-99-6492-5\_48",

language = "英语",

isbn = "9789819964918",

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

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "564--574",

editor = "Huayong Yang and Honghai Liu and Jun Zou and Zhouping Yin and Lianqing Liu and Geng Yang and Xiaoping Ouyang and Zhiyong Wang",

booktitle = "Intelligent Robotics and Applications - 16th International Conference, ICIRA 2023, Proceedings",

address = "德国",

}

Wang, F, Chen, X, Wang, X, Ren, W & Tang, Y 2023, Research on Object Detection Methods in Low-Light Conditions. in H Yang, H Liu, J Zou, Z Yin, L Liu, G Yang, X Ouyang & Z Wang (eds), Intelligent Robotics and Applications - 16th International Conference, ICIRA 2023, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 14270 LNAI, Springer Science and Business Media Deutschland GmbH, pp. 564-574, 16th International Conference on Intelligent Robotics and Applications, ICIRA 2023, Hangzhou, China, 5/07/23. https://doi.org/10.1007/978-981-99-6492-5_48

Research on Object Detection Methods in Low-Light Conditions. / Wang, Feifan; Chen, Xi’ai; Wang, Xudong et al.
Intelligent Robotics and Applications - 16th International Conference, ICIRA 2023, Proceedings. ed. / Huayong Yang; Honghai Liu; Jun Zou; Zhouping Yin; Lianqing Liu; Geng Yang; Xiaoping Ouyang; Zhiyong Wang. Springer Science and Business Media Deutschland GmbH, 2023. p. 564-574 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 14270 LNAI).

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

TY - GEN

T1 - Research on Object Detection Methods in Low-Light Conditions

AU - Wang, Feifan

AU - Chen, Xi’ai

AU - Wang, Xudong

AU - Ren, Weihong

AU - Tang, Yandong

N1 - Publisher Copyright: © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd 2023.

PY - 2023

Y1 - 2023

N2 - Low-light images are images taken in poorly illuminated environments. Such images suffer from colour distortion, loss of detail and blurriness, which seriously affects the detection accuracy of object detection tasks. In order to improve the accuracy of object detection in low-light images, we propose a low-light image object detection algorithm based on image enhancement. The algorithm is jointly trained on the input side of the YOLOv5 network in combination with an unsupervised low-light enhancement model. The training phase optimises the overall network with the loss of object detection so that the image enhancement results are more favourable for improving the object detection accuracy. In the feature extraction phase, we design a feature enhancement model based on an attention mechanism. Our algorithm is tested on the publicly available ExDark dataset and achieves a mean average precision (mAP) of 79.15%, which is a 4.25% improvement over the baseline.

AB - Low-light images are images taken in poorly illuminated environments. Such images suffer from colour distortion, loss of detail and blurriness, which seriously affects the detection accuracy of object detection tasks. In order to improve the accuracy of object detection in low-light images, we propose a low-light image object detection algorithm based on image enhancement. The algorithm is jointly trained on the input side of the YOLOv5 network in combination with an unsupervised low-light enhancement model. The training phase optimises the overall network with the loss of object detection so that the image enhancement results are more favourable for improving the object detection accuracy. In the feature extraction phase, we design a feature enhancement model based on an attention mechanism. Our algorithm is tested on the publicly available ExDark dataset and achieves a mean average precision (mAP) of 79.15%, which is a 4.25% improvement over the baseline.

KW - Attention mechanisms

KW - Deep learning

KW - Low-light enhancement

KW - Object detection

UR - https://www.scopus.com/pages/publications/85175871930

U2 - 10.1007/978-981-99-6492-5_48

DO - 10.1007/978-981-99-6492-5_48

M3 - 会议稿件

AN - SCOPUS:85175871930

SN - 9789819964918

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

SP - 564

EP - 574

BT - Intelligent Robotics and Applications - 16th International Conference, ICIRA 2023, Proceedings

A2 - Yang, Huayong

A2 - Liu, Honghai

A2 - Zou, Jun

A2 - Yin, Zhouping

A2 - Liu, Lianqing

A2 - Yang, Geng

A2 - Ouyang, Xiaoping

A2 - Wang, Zhiyong

PB - Springer Science and Business Media Deutschland GmbH

T2 - 16th International Conference on Intelligent Robotics and Applications, ICIRA 2023

Y2 - 5 July 2023 through 7 July 2023

ER -

Wang F, Chen X, Wang X, Ren W, Tang Y. Research on Object Detection Methods in Low-Light Conditions. In Yang H, Liu H, Zou J, Yin Z, Liu L, Yang G, Ouyang X, Wang Z, editors, Intelligent Robotics and Applications - 16th International Conference, ICIRA 2023, Proceedings. Springer Science and Business Media Deutschland GmbH. 2023. p. 564-574. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-981-99-6492-5_48