基于目标检测与边缘分割的输电走廊隐患预警方法

doi:10.19595/j.cnki.1000-6753.tces.250151

Abstract
Figure/Table
References
Related Citation (15)

Download: PDF (112411 KB) HTML (1 KB)
Export: BibTeX | EndNote (RIS)

Abstract As a critical carrier for power transmission, transmission lines are frequently exposed to high safety risks and recurrent accident hazards due to their characteristics of wide distribution, extensive coverage, long distances, and prolonged exposure to outdoor environments. In recent years, with the rapid development of deep learning technologies, image recognition based on deep learning has provided innovative solutions for transmission line inspection. However, challenges such as complex transmission corridor scenarios, significant variations in target scales, and low detection accuracy for mechanical external damage targets under occluded conditions often lead to false positives and missed detections. Moreover, merely identifying mechanical targets is insufficient to meet the comprehensive requirements of hazard warning, as it necessitates integrating spatial location information of transmission lines for holistic risk assessment.
To address these issues, this paper proposes a hazard detection method for transmission corridors based on edge segmentation and object detection. By incorporating a small target detection layer and an SBA module into YOLOv8, the method significantly enhances multi-scale feature representation and target localization capabilities through selective aggregation of boundary and semantic information, adaptive attention mechanisms, and bidirectional feature fusion, demonstrating superior performance in small target detection. Additionally, the adoption of re-parameterized lightweight heads and re-parameterizable convolutions reduces the number of parameters while improving parameter utilization, effectively compensating for potential accuracy loss due to lightweight design and providing a lossless optimization solution for resource-constrained devices. Furthermore, the improvement of CIoU using MPDIoU further enhances detection accuracy. This paper also employs a segmentation network for power line extraction and introduces a multi-level early warning discrimination mechanism based on pixel overlap ratio (OR) and spatial distance, aiming to improve the accuracy and reliability of hazard target detection in power field operation scenarios. By fusing the hazard target detection bounding boxes output by the object detection network with the transmission corridor regions extracted by the segmentation network, the system leverages their pixel coordinates for localization and systematically analyzes their topological relationships in pixel space. The overlap ratio (OR) between the hazard target bounding box and the transmission corridor region is defined as the core indicator for warning triggers. When an overlap exists, an OR-based warning mechanism is activated, with spatial distance further introduced as an auxiliary metric to assess the threat level of the hazard target.
The ablation experiments demonstrate that our proposed strategies achieve 72.1% mAP50, representing a 3.2 percentage points improvement over the baseline model. Comparative experiments confirm that our model outperforms YOLOv5, YOLOv7, YOLOv8, YOLOv10, and YOLOv11. Visualization results indicate that the enhanced model effectively reduces missed detections and false positives by focusing more precisely on hazard target features. The warning system evaluation proves that our segmentation-based approach for power line edge extraction provides superior foreground-background differentiation. The proposed warning method not only detects potential hazards but also accurately assesses their threat level to transmission corridors in 2D space.
Experimental results show that the proposed method can accurately identify various external damage targets in transmission corridors while quantitatively assessing the threat level of potential hazards. The excellent performance of this method in both target identification and threat analysis provides key technical support for intelligent monitoring and maintenance of modern power infrastructure, effectively promoting the innovative development of smart grid technologies.

Key words： External damage selective boundary aggregation (SBA) rep shared convolutional detection head (RSCD) safety zone minimum point distance based IoU (MPDIoU)

Received: 24 January 2025

PACS:

TM75

	Service

	E-mail this article
	Add to my bookshelf
	Add to citation manager
	E-mail Alert
	RSS
	Articles by authors
	Zhao Zhenbing
	Fu Longmei
	Pan Yitian
	Li Haopeng

Cite this article:

Zhao Zhenbing,Fu Longmei,Pan Yitian等. Early Warning Method of Transmission Corridor Hazards Based on Object Detection and Edge Segmentation[J]. Transactions of China Electrotechnical Society, 2026, 41(3): 987-998.

URL:

https://dgjsxb.ces-transaction.com/EN/10.19595/j.cnki.1000-6753.tces.250151 OR https://dgjsxb.ces-transaction.com/EN/Y2026/V41/I3/987

[1] 郑含博, 李金恒, 刘洋, 等. 基于改进YOLOv3的电力设备红外目标检测模型[J]. 电工技术学报, 2021, 36(7): 1389-1398.
Zheng Hanbo, Li Jinheng, Liu Yang, et al.Infrared object detection model for power equipment based on improved YOLOv3[J]. Transactions of China Electrotechnical Society, 2021, 36(7): 1389-1398.
[2] 宋耐超, 王瑞琦, 李明明, 等. 多自然灾害下的架空输电线路运行风险评估[J]. 电力系统保护与控制, 2021, 49(19): 65-71.
Song Naichao, Wang Ruiqi, Li Mingming, et al.Risk assessment of overhead transmission lines under multiple natural disasters[J]. Power System Protection and Control, 2021, 49(19): 65-71.
[3] Girshick R, Donahue J, Darrell T, et al.Rich feature hierarchies for accurate object detection and semantic segmentation[C]//2014 IEEE Conference on Computer Vision and Pattern Recognition, Columbus, OH, USA, 2014: 580-587.
[4] 仲林林, 胡霞, 刘柯妤. 基于改进生成对抗网络的无人机电力杆塔巡检图像异常检测[J]. 电工技术学报, 2022, 37(9): 2230-2240, 2262.
Zhong Linlin, Hu Xia, Liu Keyu.Power tower anomaly detection from unmanned aerial vehicles inspection images based on improved generative adversarial network[J]. Transactions of China Electrotechnical Society, 2022, 37(9): 2230-2240, 2262.
[5] 张烨, 李博涛, 尚景浩, 等. 基于多尺度卷积注意力机制的输电线路防振锤缺陷检测[J]. 电工技术学报, 2024, 39(11): 3522-3537.
Zhang Ye, Li Botao, Shang Jinghao, et al.Defect detection of transmission line damper based on multi-scale convolutional attention mechanism[J]. Transac-tions of China Electrotechnical Society, 2024, 39(11): 3522-3537.
[6] 姜香菊, 王瑞彤, 马彦鸿. 基于轻量级改进RT-DETR边缘部署算法的绝缘子缺陷检测[J]. 电工技术学报, 2025, 40(3): 842-854.
Jiang Xiangju, Wang Ruitong, Ma Yanhong.Insulator defect detection based on lightweight improved RT-DETR edge deployment algorithm[J]. Transactions of China Electrotechnical Society, 2025, 40(3): 842-854.
[7] 苟军年, 杜愫愫, 刘力. 基于改进掩膜区域卷积神经网络的输电线路绝缘子自爆检测[J]. 电工技术学报, 2023, 38(1): 47-59.
Xun Junnian, Du Susu, Liu Li.Transmission line insulator self-explosion detection based on improved mask region-convolutional neural network[J]. Transactions of China Electrotechnical Society, 2023, 38(1): 47-59.
[8] Zheng Hanbo, Cui Yaohui, Yang Wenqiang, et al.An infrared image detection method of substation equipment combining iresgroup structure and CenterNet[J]. IEEE Transactions on Power Delivery, 2022, 37(6): 4757-4765.
[9] 裴少通, 张行远, 胡晨龙, 等. 基于ER-YOLO算法的跨环境输电线路缺陷识别方法[J]. 电工技术学报, 2024, 39(9): 2825-2840.
Pei Shaotong, Zhang Hangyuan, Hu Chenlong, et al.The defect detection method for cross-environment power transmission line based on ER-YOLO algorithm[J]. Transactions of China Electrotechnical Society, 2024, 39(9): 2825-2840.
[10] 魏贤哲, 卢武, 赵文彬, 等. 基于改进Mask R-CNN的输电线路防外破目标检测方法研究[J]. 电力系统保护与控制, 2021, 49(23): 155-162.
Wei Xianzhe, Lu Wu, Zhao Wenbin, et al.Target detection method for external damage of a transmission line based on an improved Mask R-CNN algorithm[J]. Power System Protection and Control, 2021, 49(23): 155-162.
[11] He Kaiming, Gkioxari G, Dollár P, et al.Mask R-CNN[C]//2017 IEEE International Conference on Computer Vision (ICCV), Venice, Italy, 2017: 2980-2988.
[12] 郑含博, 胡思佳, 梁炎燊, 等. 基于YOLO-2MCS的输电线路走廊隐患目标检测方法[J]. 电工技术学报, 2024, 39(13): 4164-4175.
Zheng Hanbo, Hu Sijia, Liang Yanshen, et al.A hidden danger object detection method for transmission line corridor based on YOLO-2MCS[J]. Transactions of China Electrotechnical Society, 2024, 39(13): 4164-4175
[13] Zhao Zhenbing, Pan Yitian, Guo Guangxue, et al.YOLO-AFPN: Marrying YOLO and AFPN for external damage detection of transmission lines[J]. IET Generation, Transmission & Distribution, 2024, 18(9): 1935-1946.
[14] Gugel H, Ekisheva S, Lauby M, et al.Vegetation-related outages on transmission lines in North America[C]//2018 IEEE Power & Energy Society General Meeting (PESGM), Portland, OR, USA, 2018: 1-5.
[15] Nardinocchi C, Balsi M, Esposito S.Fully automatic point cloud analysis for powerline corridor mapping[J]. IEEE Transactions on Geoscience and Remote Sensing, 2020, 58(12): 8637-8648.
[16] Chen Yangyu, Lin Jiayuan, Liao Xiaohan.Early detection of tree encroachment in high voltage powerline corridor using growth model and UAV-borne LiDAR[J]. International Journal of Applied Earth Observation and Geoinformation, 2022, 108: 102740.
[17] Li Jinheng, Zheng Hanbo, Cui Zhimei, et al.Intelligent detection method with 3D ranging for external force damage monitoring of power transmission lines[J]. Applied Energy, 2024, 374: 123983.
[18] Wang Leixiong, Wang Bo, Wang Shulong, et al.An effective method for sensing power safety distance based on monocular vision depth estimation[J]. International Transactions on Electrical Energy Systems, 2023(1): 8480342.
[19] Ma Fuqi, Liu Heng, Wang Jiaxun, et al.A refined identification method for the hidden dangers of external damage in transmission lines based on the generation of a vanishing point-driven effective region[J]. Processes, 2024, 12(9): 1904.
[20] 刘波. 面向复杂输电线路区域的外破隐患识别[D]. 济南: 山东大学, 2022.
Liu Bo.Wire region-oriented hazard detection in the complex transmission line[D]. Jinan: Shandong University, 2022.
[21] Varghese R, Sambath M.YOLOv8: a novel object detection algorithm with enhanced performance and robustness[C]//2024 International Conference on Advances in Data Engineering and Intelligent Computing Systems (ADICS), Chennai, India, 2024: 1-6.
[22] Tang Feilong, Xu Zhongxing, Huang Qiming, et al.DuAT: dual-aggregation transformer network for medical image segmentation[M]//Liu Q, Wang H, Ma Z, et al. Lecture Notes in Computer Science. Singapore: Springer Nature Singapore, 2023: 343-356.
[23] Chen Jieneng, Lu Yongyi, Yu Qihang, et al. TransUNet: transformers make strong encoders for medical image segmentation[J/OL]. ArXiv, 2021: 2102.04306[2025-01-10]. https://arxiv.org/abs/2102.04306v1.
[24] Wang Ao, Chen Hui, Liu Lihao, et al. YOLOv10: real-time end-to-end object detection[J/OL]. ArXiv, 2024: 2405.14458[2025-01-10]. https://arxiv.org/abs/2405.14458v2.
[25] Khanam R, Hussain M. YOLOv11: an overview of the key architectural enhancements[J/OL]. ArXiv, 2024: 2410.17725[2025-01-10]. https://arxiv.org/abs/2410.17725v1.
[26] Chen L C, Zhu Yukun, Papandreou G, et al.Encoder-decoder with atrous separable convolution for semantic image segmentation[M]//Ferrari V, Hebert M, Sminchisescu C, et al. Lecture Notes in Computer Science. Cham: Springer International Publishing, 2018: 833-851.
[27] Yu Changqian, Gao Changxin, Wang Jingbo, et al.BiSeNet V2: bilateral network with guided aggregation for real-time semantic segmentation[J]. International Journal of Computer Vision, 2021, 129(11): 3051-3068.
[28] Zhao Hengshuang, Shi Jianping, Qi Xiaojuan, et al.Pyramid scene parsing network[C]//2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Honolulu, HI, USA, 2017: 6230-6239.
[29] Yin M, Yao Z, Cao Y, et al.Disentangled non-local neural networks[C]//Computer Vision-ECCV 2020: 16th European Conference, Glasgow, UK, 2020: 191-207.