Research on Image Recognition Algorithms for Foreign Object Intrusion in Transmission Lines

Fuguo Chen; Mingzhe Liu; Huan Yuan; Aijun Yang; Xiaohua Wang

doi:10.1109/CEEPE64987.2025.11033794

Research on Image Recognition Algorithms for Foreign Object Intrusion in Transmission Lines

Fuguo Chen
, Mingzhe Liu
, Huan Yuan
, Aijun Yang
, Xiaohua Wang

School of Electrical Engineering

Xi'an Jiaotong University
Ltd.

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

Abstract

With the increasing frequency of external damages to transmission lines, traditional manual monitoring methods consume significant human resources. This paper proposes a computer vision-based image recognition algorithm for foreign object intrusion in transmission lines. First, a transmission line recognition algorithm based on the instance segmentation model LaneNet is proposed, achieving an accuracy of 90.85%. Second, a foreign object detection algorithm based on the DenseBox model is introduced, with an accuracy of 92.25%. Finally, a monocular vision-based 3D object localization algorithm is proposed, with experimental results showing an error below 3.11%. The proposed algorithm offers an effective solution for the automation of transmission line monitoring, enhancing the automation and intelligence of power grid operation and maintenance.

Original language	English
Title of host publication	2025 8th International Conference on Energy, Electrical and Power Engineering, CEEPE 2025
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	268-274
Number of pages	7
ISBN (Electronic)	9798331521844
DOIs	https://doi.org/10.1109/CEEPE64987.2025.11033794
State	Published - 2025
Event	8th International Conference on Energy, Electrical and Power Engineering, CEEPE 2025 - Wuxi, China Duration: 25 Apr 2025 → 27 Apr 2025

Publication series

Name	2025 8th International Conference on Energy, Electrical and Power Engineering, CEEPE 2025

Conference

Conference	8th International Conference on Energy, Electrical and Power Engineering, CEEPE 2025
Country/Territory	China
City	Wuxi
Period	25/04/25 → 27/04/25

UN SDGs

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

SDG 7 Affordable and Clean Energy

Keywords

image recognition
instance segmentation
monocular vision
object detection
transmission line

Access to Document

10.1109/CEEPE64987.2025.11033794

Cite this

Chen, F., Liu, M., Yuan, H., Yang, A., & Wang, X. (2025). Research on Image Recognition Algorithms for Foreign Object Intrusion in Transmission Lines. In 2025 8th International Conference on Energy, Electrical and Power Engineering, CEEPE 2025 (pp. 268-274). (2025 8th International Conference on Energy, Electrical and Power Engineering, CEEPE 2025). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/CEEPE64987.2025.11033794

Chen, Fuguo ; Liu, Mingzhe ; Yuan, Huan et al. / Research on Image Recognition Algorithms for Foreign Object Intrusion in Transmission Lines. 2025 8th International Conference on Energy, Electrical and Power Engineering, CEEPE 2025. Institute of Electrical and Electronics Engineers Inc., 2025. pp. 268-274 (2025 8th International Conference on Energy, Electrical and Power Engineering, CEEPE 2025).

@inproceedings{afacfa48ae9e4dc186b130d0abb8c915,

title = "Research on Image Recognition Algorithms for Foreign Object Intrusion in Transmission Lines",

abstract = "With the increasing frequency of external damages to transmission lines, traditional manual monitoring methods consume significant human resources. This paper proposes a computer vision-based image recognition algorithm for foreign object intrusion in transmission lines. First, a transmission line recognition algorithm based on the instance segmentation model LaneNet is proposed, achieving an accuracy of 90.85\%. Second, a foreign object detection algorithm based on the DenseBox model is introduced, with an accuracy of 92.25\%. Finally, a monocular vision-based 3D object localization algorithm is proposed, with experimental results showing an error below 3.11\%. The proposed algorithm offers an effective solution for the automation of transmission line monitoring, enhancing the automation and intelligence of power grid operation and maintenance.",

keywords = "image recognition, instance segmentation, monocular vision, object detection, transmission line",

author = "Fuguo Chen and Mingzhe Liu and Huan Yuan and Aijun Yang and Xiaohua Wang",

note = "Publisher Copyright: {\textcopyright} 2025 IEEE.; 8th International Conference on Energy, Electrical and Power Engineering, CEEPE 2025 ; Conference date: 25-04-2025 Through 27-04-2025",

year = "2025",

doi = "10.1109/CEEPE64987.2025.11033794",

language = "英语",

series = "2025 8th International Conference on Energy, Electrical and Power Engineering, CEEPE 2025",

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

pages = "268--274",

booktitle = "2025 8th International Conference on Energy, Electrical and Power Engineering, CEEPE 2025",

}

Chen, F, Liu, M, Yuan, H , Yang, A & Wang, X 2025, Research on Image Recognition Algorithms for Foreign Object Intrusion in Transmission Lines. in 2025 8th International Conference on Energy, Electrical and Power Engineering, CEEPE 2025. 2025 8th International Conference on Energy, Electrical and Power Engineering, CEEPE 2025, Institute of Electrical and Electronics Engineers Inc., pp. 268-274, 8th International Conference on Energy, Electrical and Power Engineering, CEEPE 2025, Wuxi, China, 25/04/25. https://doi.org/10.1109/CEEPE64987.2025.11033794

Research on Image Recognition Algorithms for Foreign Object Intrusion in Transmission Lines. / Chen, Fuguo; Liu, Mingzhe; Yuan, Huan et al.
2025 8th International Conference on Energy, Electrical and Power Engineering, CEEPE 2025. Institute of Electrical and Electronics Engineers Inc., 2025. p. 268-274 (2025 8th International Conference on Energy, Electrical and Power Engineering, CEEPE 2025).

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

TY - GEN

T1 - Research on Image Recognition Algorithms for Foreign Object Intrusion in Transmission Lines

AU - Chen, Fuguo

AU - Liu, Mingzhe

AU - Yuan, Huan

AU - Yang, Aijun

AU - Wang, Xiaohua

PY - 2025

Y1 - 2025

N2 - With the increasing frequency of external damages to transmission lines, traditional manual monitoring methods consume significant human resources. This paper proposes a computer vision-based image recognition algorithm for foreign object intrusion in transmission lines. First, a transmission line recognition algorithm based on the instance segmentation model LaneNet is proposed, achieving an accuracy of 90.85%. Second, a foreign object detection algorithm based on the DenseBox model is introduced, with an accuracy of 92.25%. Finally, a monocular vision-based 3D object localization algorithm is proposed, with experimental results showing an error below 3.11%. The proposed algorithm offers an effective solution for the automation of transmission line monitoring, enhancing the automation and intelligence of power grid operation and maintenance.

AB - With the increasing frequency of external damages to transmission lines, traditional manual monitoring methods consume significant human resources. This paper proposes a computer vision-based image recognition algorithm for foreign object intrusion in transmission lines. First, a transmission line recognition algorithm based on the instance segmentation model LaneNet is proposed, achieving an accuracy of 90.85%. Second, a foreign object detection algorithm based on the DenseBox model is introduced, with an accuracy of 92.25%. Finally, a monocular vision-based 3D object localization algorithm is proposed, with experimental results showing an error below 3.11%. The proposed algorithm offers an effective solution for the automation of transmission line monitoring, enhancing the automation and intelligence of power grid operation and maintenance.

KW - image recognition

KW - instance segmentation

KW - monocular vision

KW - object detection

KW - transmission line

UR - https://www.scopus.com/pages/publications/105009597095

U2 - 10.1109/CEEPE64987.2025.11033794

DO - 10.1109/CEEPE64987.2025.11033794

M3 - 会议稿件

AN - SCOPUS:105009597095

T3 - 2025 8th International Conference on Energy, Electrical and Power Engineering, CEEPE 2025

SP - 268

EP - 274

BT - 2025 8th International Conference on Energy, Electrical and Power Engineering, CEEPE 2025

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 8th International Conference on Energy, Electrical and Power Engineering, CEEPE 2025

Y2 - 25 April 2025 through 27 April 2025

ER -

Chen F, Liu M, Yuan H , Yang A , Wang X. Research on Image Recognition Algorithms for Foreign Object Intrusion in Transmission Lines. In 2025 8th International Conference on Energy, Electrical and Power Engineering, CEEPE 2025. Institute of Electrical and Electronics Engineers Inc. 2025. p. 268-274. (2025 8th International Conference on Energy, Electrical and Power Engineering, CEEPE 2025). doi: 10.1109/CEEPE64987.2025.11033794

Research on Image Recognition Algorithms for Foreign Object Intrusion in Transmission Lines

Abstract

Publication series

Conference

UN SDGs

Keywords

Access to Document

Other files and links

Fingerprint

Cite this