Cascade detection method for insulator defects in distribution lines based on improved YOLOv8

doi:10.11996/JG.j.2095-302X.2025010001

Abstract

Abstract:

To address the issues of complex and dynamic backgrounds due to safety constraints, irregular shapes of insulator defects, indistinct defect features, and difficulty in capturing defect information during aerial photography of power distribution lines using unmanned aerial vehicles, a cascaded detection method for insulator defects in power distribution lines based on an improved YOLOv8 was proposed. In the first stage, the YOLOv8 model automatically extracted images of insulator components, providing accurate inputs for the second stage of insulator defect detection and eliminating the influence of redundant background information. In the second stage, the ConvNeXt V2 backbone network was utilized to enhance the model's ability to recognize irregularly shaped targets and improve its feature extraction capabilities. By incorporating the edge knowledge fusion module into the feature fusion process, precise extraction of defect edge information was achieved. Furthermore, an adaptive shape IoU enhancement method was designed, adopting an adaptive training sample selection strategy to optimize the ratio of positive and negative samples. Additionally, the Shape-IoU loss function was employed, considering the inherent attributes of bounding box regression samples such as shape and scale, enabling the model to focus on essential target features, thereby improving the detection accuracy and robustness by reducing missed and false detections. Experimental results demonstrated that the proposed cascaded detection method for insulator defects in power distribution lines based on the improved YOLOv8 achieved a 17.3% increase in average precision compared to baseline models, effectively enhancing the accuracy of insulator defect detection in power distribution lines and providing robust technical support for the safe maintenance of power systems.

Key words: distribution lines, insulator defect detection, YOLOv8, ConvNeXt V2, edge knowledge fusion, adaptive shape IoU enhancement

CLC Number:

ZHAO Zhenbing, HAN Yu, TANG Chenkang. Cascade detection method for insulator defects in distribution lines based on improved YOLOv8[J]. Journal of Graphics, 2025, 46(1): 1-12.

Figures/Tables 12

References 37

[1]	盛戈皞, 钱勇, 罗林根, 等. 面向新型电力系统的电力设备运行维护关键技术及其应用展望[J]. 高电压技术, 2021, 47(9): 3072-3084.
	SHENG G H, QIAN Y, LUO L G, et al. Key technologies and application prospects for operation and maintenance of power equipment in new type power system[J]. High Voltage Engineering, 2021, 47(9): 3072-3084 (in Chinese).
[2]	林聚财, 韩军, 陈舫明, 等. 基于彩色图像的玻璃绝缘子缺陷诊断[J]. 电网技术, 2011, 35(1): 127-133.
	LIN J C, HAN J, CHEN F M, et al. Defects detection of glass insulator based on color image[J]. Power System Technology, 2011, 35(1): 127-133 (in Chinese).
[3]	张少平, 杨忠, 黄宵宁, 等. 航拍图像中玻璃绝缘子自爆缺陷的检测及定位[J]. 太赫兹科学与电子信息学报, 2013, 11(4): 609-613.
	ZHANG S P, YANG Z, HUANG X N, et al. Defects detection and positioning for glass insulator from aerial images[J]. Journal of Terahertz Science and Electronic Information Technology, 2013, 11(4): 609-613 (in Chinese).
[4]	梅欣, 陈江, 王书友. 基于形态学的复合绝缘子憎水性图像检测技术[J]. 高电压技术, 2013, 39(12): 3088-3094.
	MEI X, CHEN J, WANG S Y. Composite insulator hydrophobicity image detection method based on morphology[J]. High Voltage Engineering, 2013, 39(12): 3088-3094 (in Chinese).
[5]	NI L Y, MA Y J, LIN Q, et al. Research on insulator defect detection method based on image processing and watershed algorithm[C]// 2021 International Conference on Advanced Electrical Equipment and Reliable Operation. New York: IEEE Press, 2021: 1-6.
[6]	韩俊, 袁小平, 王准, 等. 基于YOLOv5s的无人机密集小目标检测算法[J]. 浙江大学学报(工学版), 2023, 57(6): 1224-1233.
	HAN J, YUAN X P, WANG Z, et al. UAV dense small target detection algorithm based on YOLOv5s[J]. Journal of Zhejiang University (Engineering Science), 2023, 57(6): 1224-1233 (in Chinese).
[7]	SHARMA K U, THAKUR N V. A review and an approach for object detection in images[J]. International Journal of Computational Vision and Robotics, 2017, 7(1/2): 196-237.
[8]	LIU W, ANGUELOV D, ERHAN D, et al. SSD: single shot MultiBox detector[C]// The 14th European Conference on Computer Vision-ECCV 2016. Cham: Springer, 2016: 21-37.
[9]	REN S Q, HE K M, GIRSHICK R, et al. Faster R-CNN: towards real-time object detection with region proposal networks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(6): 1137-1149. DOI PMID
[10]	REDMON J, DIVVALA S, GIRSHICK R, et al. You only look once: unified, real-time object detection[C]// 2016 IEEE Conference on Computer Vision and Pattern Recognition. New York: IEEE Press, 2016: 779-788.
[11]	朱泓宇, 程换新, 骆晓玲. 改进YOLOv8网络在绝缘子缺陷检测中的应用[J]. 水电能源科学, 2024, 42(5): 183-187.
	ZHU H Y, CHENG H X, LUO X L. Application of improved YOLOv8 network in insulator defect detection[J]. Water Resources and Power, 2024, 42(5): 183-187 (in Chinese).
[12]	王思, 熊惠敏, 胡蕾, 等. 基于改进YOLOv7的输电线路绝缘子设备及其缺陷检测算法[J]. 江西电力, 2024, 48(1): 1-6, 20.
	WANG S, XIONG H M, HU L, et al. Detection algorithm for transmission line insulator equipment and its defects based on improved YOLOv7[J]. Jiangxi Electric Power, 2024, 48(1): 1-6, 20 (in Chinese).
[13]	卞建鹏, 朱泽明, 陈璇, 等. 基于Res-CapsNet与改进YOLOv4的绝缘子破损识别与定位[J/OL]. (2024-04-01) [2024-06-20]. http://kns.cnki.net/kcms/detail/61.1129.TM.20240330.1344.006.html.
	BIAN J P, ZHU Z M, CHEN X, et al. Insulator damage identification and location based on Res-CapsNet and improved YOLOv4[J/OL]. (2024-04-01) [2024-09-20]. http://kns.cnki.net/kcms/detail/61.1129.TM.20240330.1344.006.html (in Chinese).
[14]	WANG S Q, LIU Y F, QING Y, et al. Detection of insulator defects with improved ResNeSt and region proposal network[J]. IEEE Access, 2020, 8: 184841-184850.
[15]	曾业战, 段志超, 郭彦东, 等. 基于ScSGB-RCNN网络的输电线路航拍绝缘子目标检测[J]. 电瓷避雷器, 2024(1): 161-169.
	ZENG Y Z, DUAN Z C, GUO Y D, et al. Target detection for aerial imaging insulators of transmission lines based on ScSGB-RCNN network[J]. Insulators and Surge Arresters, 2024(1): 161-169 (in Chinese).
[16]	翟永杰, 杨旭, 赵振兵, 等. 融合共现推理的Faster R-CNN输电线路金具检测[J]. 智能系统学报, 2021, 16(2): 237-246.
	ZHAI Y J, YANG X, ZHAO Z B, et al. Integrating co-occurrence reasoning for Faster R-CNN transmission line fitting detection[J]. CAAI Transactions on Intelligent Systems, 2021, 16(2): 237-246 (in Chinese).
[17]	赵振兵, 江爱雪, 戚银城, 等. 嵌入遮挡关系模块的SSD模型的输电线路图像金具检测[J]. 智能系统学报, 2020, 15(4): 656-662.
	ZHAO Z B, JIANG A X, QI Y C, et al. Fittings detection in transmission line images with SSD model embedded occlusion relation module[J]. CAAI Transactions on Intelligent Systems, 2020, 15(4): 656-662 (in Chinese).
[18]	李刚, 蔡泽浩, 孙华勋, 等. 基于改进YOLOv8与语义知识融合的金具缺陷检测方法研究[J]. 图学学报, 2024, 45(5): 979-986. DOI
	LI G, CAI Z H, SUN H X, et al. Research on defect detection of transmission line fittings based on improved YOLOv8 and semantic knowledge fusion[J]. Journal of Graphics, 2024, 45(5): 979-986 (in Chinese). DOI
[19]	翟永杰, 赵晓瑜, 王璐瑶, 等. IDD-YOLOv7: 一种用于输电线路绝缘子多缺陷的轻量化检测方法[J]. 图学学报, 2024, 45(1): 90-101. DOI
	ZHAI Y J, ZHAO X Y, WANG L Y, et al. IDD-YOLOv7: a lightweight method for multiple defect detection of insulators in transmission lines[J]. Journal of Graphics, 2024, 45(1): 90-101 (in Chinese). DOI
[20]	曾业战, 郭彦东, 段志超, 等. 基于SEPyconv-DETR的无人机航拍绝缘子检测算法[J]. 电工技术, 2024(3): 27-31.
	ZENG Y Z, GUO Y D, DUAN Z C, et al. SEPyconv-DETR- based Drone-assisted insulator detection algorithm[J]. Electric Engineering, 2024(3): 27-31 (in Chinese).
[21]	杨永娇, 唐亮亮, 郑勇. 改进全卷积网络优化绝缘子缺陷检测[J]. 机械设计与制造, 2021(3): 177-180.
	YANG Y J, TANG L L, ZHENG Y. Insulator defect detection optimized by improved full convolution network[J]. Machinery Design & Manufacture, 2021(3): 177-180 (in Chinese).
[22]	ZHANG X T, ZHANG Y Y, LIU J F, et al. InsuDet: a fault detection method for insulators of overhead transmission lines using convolutional neural networks[J]. IEEE Transactions on Instrumentation and Measurement, 2021, 70: 5018512.
[23]	王道累, 孙嘉珺, 张天宇, 等. 基于改进生成对抗网络的玻璃绝缘子自爆缺陷检测方法[J]. 高电压技术, 2022, 48(3): 1096-1103.
	WANG D L, SUN J J, ZHANG T Y, et al. Self-explosion defect detection method of glass insulator based on improved generative adversarial network[J]. High Voltage Engineering, 2022, 48(3): 1096-1103 (in Chinese).
[24]	颜宏文, 万俊杰, 潘志敏, 等. 基于改进YOLOv5-LITE轻量级的配电组件缺陷识别[J]. 高电压技术, 2024, 50(5): 1855-1864.
	YAN H W, WAN J J, PAN Z M, et al. Defect identification of distribution components based on improved YOLOv5-LITE lightweight[J]. High Voltage Engineering, 2024, 50(5): 1855-1864 (in Chinese).
[25]	吕继军, 陈涛, 李梦琪, 等. 基于计算机视觉的配电线路绝缘子合格检测[J]. 光源与照明, 2023(8): 108-110.
	LV J J, CHEN T, LI M Q, et al. Qualification detection of distribution line insulators based on computer vision[J]. Light & Lighting, 2023(8): 108-110 (in Chinese).
[26]	李利荣, 张云良, 陈鹏, 等. 基于上下文信息增强与特征细化的绝缘子破损检测方法[J]. 高电压技术, 2023, 49(8): 3405-3414.
	LI L R, ZHANG Y L, CHEN P, et al. Detection method of insulator breakage based on context augmentation and feature refinement[J]. High Voltage Engineering, 2023, 49(8): 3405-3414 (in Chinese).
[27]	LU Z X, LI Y, SHUANG F, et al. InsDef: few-shot learning-based insulator defect detection algorithm with a dual-guide attention mechanism and multiple label consistency constraints[J]. IEEE Transactions on Power Delivery, 2023, 38(6): 4166-4178.
[28]	WOO S, DEBNATH S, HU R H, et al. ConvNeXt V2: co-designing and scaling ConvNets with masked autoencoders[C]// 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition. New York: IEEE Press, 2023: 16133-16142.
[29]	ZHANG S F, CHI C, YAO Y Q, et al. Bridging the gap between anchor-based and anchor-free detection via adaptive training sample selection[C]// 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. New York: IEEE Press, 2020: 9756-9765.
[30]	ZHANG H, ZHANG S J. Shape-IoU: more accurate metric considering bounding box shape and scale[EB/OL]. [2024-05-20]. https://arxiv.org/abs/2312.17663.
[31]	LI C Y, LI L L, JIANG H L, et al. YOLOv6: a single-stage object detection framework for industrial applications[EB/OL]. [2024-05-20]. https://arxiv.org/abs/2209.02976.
[32]	WANG C Y, BOCHKOVSKIY A, LIAO H Y M. YOLOv7: trainable bag-of-freebies sets new state-of-the-art for real-time object detectors[C]// 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition. New York: IEEE Press, 2023: 7464-7475.
[33]	WANG C Y, YEH I H, LIAO H Y M. YOLOv9: learning what you want to learn using programmable gradient information[EB/OL]. [2024-06-20]. https://doi.org/10.48550/arXiv.2402.13616.
[34]	LIU X Y, PENG H W, ZHENG N X, et al. EfficientViT: memory efficient vision transformer with cascaded group attention[C]// 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition. New York: IEEE Press, 2023: 14420-14430.
[35]	CHEN J R, KAO S H, HE H, et al. Run, don't walk: chasing higher FLOPS for faster neural networks[C]// 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition. New York: IEEE Press, 2023: 12021-12031.
[36]	CHEN H T, WANG Y H, GUO J Y, et al. VanillaNet: the power of minimalism in deep learning[EB/OL]. [2024-06-20]. https://proceedings.neurips.cc/paper_files/paper/2023/file/16336d94a5ffca8de019087ab7fe403f-Supplemental-Conference.pdf.
[37]	LI Y X, LI X, DAI Y, et al. LSKNet: a foundation lightweight backbone for remote sensing[EB/OL]. [2024-05-20]. https://link.springer.com/article/10.1007/s11263-024-02247-9.

名称	型号
操作系统	Ubuntu16.04.1
CPU	E5-2620 v4
GPU	TITAN Xp
CUDA	10.2
cuDNN	7.6.5
Python	3.8
深度学习框架	PyTorch 1.12.0

名称	型号
操作系统	Ubuntu16.04.1
CPU	E5-2620 v4
GPU	TITAN Xp
CUDA	10.2
cuDNN	7.6.5
Python	3.8
深度学习框架	PyTorch 1.12.0

目标类别	绝缘子数据集			绝缘子缺陷数据集
目标类别	悬式绝缘子	柱式绝缘子	针式绝缘子	损伤绝缘子	污秽绝缘子	正常绝缘子
样本数量/个	3 831	5 333	2 489	430	1 160	463
标签名称	x_jyz	zh_jyz	zs_jyz	sunshang	wuhui	zq

目标类别	绝缘子数据集			绝缘子缺陷数据集
目标类别	悬式绝缘子	柱式绝缘子	针式绝缘子	损伤绝缘子	污秽绝缘子	正常绝缘子
样本数量/个	3 831	5 333	2 489	430	1 160	463
标签名称	x_jyz	zh_jyz	zs_jyz	sunshang	wuhui	zq

主干网络	P/%	R/%	mAP50/%	mAP50:95/%	Parameters/M	GFLOPs/%
ConvNeXt V2^[28]	69.8	67.2	75.2	67.1	5.7	14.1
EfficientViT^[34]	53.3	67.1	61.4	48.7	4.0	9.4
FasterNet^[35]	59.4	70.0	65.6	53.7	4.2	10.7
Timm	61.5	71.5	68.7	58.7	13.3	35.1
VanillaNet^[36]	61.8	66.6	69.6	58.3	23.9	96.7
LSKNet^[37]	53.1	74.3	67.6	56.5	5.9	19.7