[1]廖晓辉,谢子晨,辛忠良,等.基于轻量化YOLOv5的电气设备外部缺陷检测[J].郑州大学学报(工学版),2024,45(04):117-124.[doi:10.13705/ j.issn.1671-6833.2024.04.010]
 LIAO Xiaohui,XIE Zichen,XIN Zhongliang,et al.Electrical Equipment External Defect Detection Based on Lightweight YOLOv5[J].Journal of Zhengzhou University (Engineering Science),2024,45(04):117-124.[doi:10.13705/ j.issn.1671-6833.2024.04.010]
点击复制

基于轻量化YOLOv5的电气设备外部缺陷检测()
分享到:

《郑州大学学报(工学版)》[ISSN:1671-6833/CN:41-1339/T]

卷:
45
期数:
2024年04期
页码:
117-124
栏目:
出版日期:
2024-06-16

文章信息/Info

Title:
Electrical Equipment External Defect Detection Based on Lightweight YOLOv5
文章编号:
1671-6833(2024)04-0117-08
作者:
廖晓辉1 谢子晨1 辛忠良2 陈 怡1 叶梁劲1
1.郑州大学 电气与信息工程学院,河南 郑州 450001;2.国网郑州供电公司,河南 郑州 450007
Author(s):
LIAO Xiaohui1 XIE Zichen1 XIN Zhongliang2 CHEN Yi1 YE Liangjin1
1.School of Electrical and Information Engineering, Zhengzhou University, Zhengzhou 450001, China; 2.State Grid Zhengzhou Pow er Supply Company, Zhengzhou 450007, China
关键词:
缺陷检测 电气设备 轻量化YOLOv5 EfficientViT网络 SimAM注意力 Soft-NMS结构
Keywords:
defect detection electrical equipment lightweight YOLOv5 EfficientViT network SimAM attention Soft-NMS structure
分类号:
TP391.4TM63
DOI:
10.13705/ j.issn.1671-6833.2024.04.010
文献标志码:
A
摘要:
为了提高变电站电气设备外部缺陷实时检测的精度,同时让检测模型更加轻量化,提出了一种基于轻量 化YOLOv5的电气设备外部缺陷检测方法。首先,构建电气设备外部缺陷图像数据集并进行数据增强处理。其 次,采用3种优化策略对原YOLOv5进行改进:通过引入EfficientViT网络改进算法主干网络,减少模型参数量,并 在算法Neck部分中加入SimAM无参数注意力机制来提高变电站复杂背景下的识别精度,同时采用Soft-NMS模块 来改进检测框筛选方式,避免出现缺陷漏检现象。最后,通过消融实验进行验证。结果表明:轻量化后的电气设备 外部缺陷检测模型mAP值稳定在86.4%,与原模型相比提高了1.2百分点,模型参数量减少了20%,计算量减少了 38%,模型大小为11 MB,比原模型减少了19.7%。改进后的模型能够满足设备外部缺陷实时检测的要求,可以实 现模型的轻量化部署。
Abstract:
In order to improve the accuracy of real-time detection of external defects of electrical equipment in sub stations and make the detection model more lightweight, a lightweight YOLOv5 based external defect detection method for electrical equipment was proposed. Firstly, the external defect image dataset of electrical equipment was constructed and processed by data enhancement. Secondly, three optimization strategies were used to improve the original YOLOv5. The EfficientViT network was introduced to improve the backbone network of the algorithm to re duce the number of model parameters, and the SimAM parameter-free attention mechanism was added to the Neck part of the algorithm to improve the recognition accuracy with the complex background of the substation. At the same time, the Soft-NMS module was used to improve the screening method of the detection box to avoid the phe nomenon of defect missed detection. Finally, verified by ablation test, the mAP value of the lightweight external de fect detection model of electrical equipment was stable at 86.4%, which was 1.2 percentage points higher than that of the original model, the number of model parameters were reduced by 20%, the calculation amount was reduced by 38%, and the model size was 11 MB, which was 19.7% lower than that of the original model. The improved model could meet the requirements of real-time detection of external defects of equipment.

参考文献/References:

[1]齐冬莲, 韩译锋, 周自强, 等. 基于视频图像的输变 电设备外部缺陷检测技术及其应用现状[J]. 电子与 信息学报, 2022, 44(11): 3709-3720. 

QI D L, HAN Y F, ZHOU Z Q, et al. Review of defect detection technology of power equipment based on video images[J]. Journal of Electronics & Information Technol ogy, 2022, 44(11): 3709-3720. 
[2] 何国立, 齐冬莲, 闫云凤. 一种基于关键点检测和注 意力机制的违规着装识别算法及其应用[J]. 中国电 机工程学报, 2022, 42(5): 1826-1837. 
HE G L, QI D L, YAN Y F. An illegal dress recognition algorithm based on key-point detection and attention mechanism and its application[J]. Proceedings of the CSEE, 2022, 42(5): 1826-1837. 
[3] GIRSHICK R. Fast R-CNN[C]∥2015 IEEE Internation al Conference on Computer Vision (ICCV). Piscataway: IEEE, 2015: 1440-1448. 
[4] 李文璞, 谢可, 廖逍, 等. 基于Faster RCNN变电设备 红外图像缺陷识别方法[J]. 南方电网技术, 2019, 13 (12): 79-84. 
LI W P, XIE K, LIAO X, et al. Intelligent diagnosis method of infrared image for transformer equipment based on improved Faster RCNN[J]. Southern Power System Technology, 2019, 13(12): 79-84. 
[5] 林刚, 王波, 彭辉, 等. 基于改进Faster-RCNN的输电 线巡检图像多目标检测及定位[J]. 电力自动化设备, 2019, 39(5): 213-218. 
LIN G, WANG B, PENG H, et al. Multi-target detection and location of transmission line inspection image based on improved Faster-RCNN[J]. Electric Power Automa tion Equipment, 2019, 39(5): 213-218. 
[6] 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 Recogni tion (CVPR). Piscataway: IEEE, 2016: 779-788. 
[7] 朱惠玲, 牛哲文, 黄克灿, 等. 基于单阶段目标检测 算法的变电设备红外图像目标识别及定位[J]. 电力 自动化设备, 2021, 41(8): 217-224. 
ZHU H L, NIU Z W, HUANG K C, et al. Identification and location of infrared image for substation equipment based on single-stage object detection algorithm[J]. Electric Power Automation Equipment, 2021, 41(8): 217-224. 
[8] 陈婷, 周旻, 韩勤, 等. 基于改进YOLOv4的变电站缺 陷检测[J]. 计算机系统应用, 2022, 31(6): 245-251. 
CHEN T, ZHOU M, HAN Q, et al. Defect detection for substation based on improved YOLOv4[J]. Computer Systems and Applications, 2022, 31(6): 245-251. 
[9] 王佰川, 王聪. 基于改进YOLOv4的配电线路绝缘子 与避雷器快速检测研究[J]. 电瓷避雷器, 2023(3): 166-174. 
WANG B C, WANG C. Rapid detection of insulators and arrester on distribution lines based on improved YOLOv4 [J]. Insulators and Surge Arresters, 2023(3): 166-174. 
[10]肖粲俊, 潘睿志, 李超, 等. 基于改进YOLOv5s绝缘 子缺陷检测技术研究[J]. 电子测量技术, 2022, 45 (24): 137-144. 
XIAO C J, PAN R Z, LI C, et al. Research on defect detection technology based on improved YOLOv5s insula tor[J]. Electronic Measurement Technology, 2022, 45 (24): 137-144. 
[11] CAI H, LI J Y, HU M Y, et al. EfficientViT: multi scale linear attention for high-resolution dense prediction [EB/OL]. (2022-05-29)[2024-02-10]. http:∥arx iv.org/abs/2205.14756. 
[12] YANG L X, ZHANG R Y, LI L D, et al. SimAM: a simple, parameter-free attention module for convolutional neural networks[C]∥International Conference on Ma chine Learning. New York: ACM, 2022:11863-11874. 
[13] BODLA N, SINGH B, CHELLAPPA R, et al. Soft NMS—improving object detection with one line of code [C]∥2017 IEEE International Conference on Computer Vision (ICCV). Piscataway: IEEE, 2017: 5562-5570. 
[14]郝帅, 杨磊, 马旭, 等. 基于注意力机制与跨尺度特 征融合的YOLOv5输电线路故障检测[J]. 中国电机 工程学报, 2023, 43(6): 2319-2331. 
HAO S, YANG L, MA X, et al. YOLOv5 transmission line fault detection based on attention mechanism and cross-scale feature fusion[J]. Proceedings of the CSEE, 2023, 43(6): 2319-2331. 
[15]曾耀, 高法钦. 基于改进YOLOv5的电子元件表面缺 陷检测算法[J]. 浙江大学学报(工学版), 2023, 57 (3): 455-465. 
ZENG Y, GAO F Q. Surface defect detection algorithm of electronic components based on improved YOLOv5[J]. Journal of Zhejiang University (Engineering Science), 2023, 57(3): 455-465. 
[16] LIU X Y, PENG H W, ZHENG N X, et al. Effi cientViT: memory efficient vision transformer with casca ded group attention[EB/OL]. (2023-05-11)[2024-02-10]. http:∥arxiv.org/abs/2305.07027. 
[17]张森, 万吉林, 王慧芳, 等. 基于注意力机制的卷积 神经网络指针式仪表图像读数识别方法[J]. 电力自 动化设备, 2022, 42(4): 218-224. 
ZHANG S, WAN J L, WANG H F, et al. Convolutional neural network based on attention mechanism for reading recognition of pointer-type meter images[J]. Electric Power Automation Equipment, 2022, 42(4): 218-224. 
[18]宁纪锋, 林靖雅, 杨蜀秦, 等. 基于改进YOLO v5s的 奶山羊面部识别方法[J]. 农业机械学报, 2023, 54 (4): 331-337. 
NING J F, LIN J Y, YANG S Q, et al. Face recognition method of dairy goat based on improved YOLO v5s[J]. Transactions of the Chinese Society for Agricultural Ma chinery, 2023, 54(4): 331-337. 
[19]高新阳, 魏晟, 温志庆, 等. 改进YOLOv5轻量级网 络的柑橘检测方法[J]. 计算机工程与应用, 2023, 59 (11): 212-221. 
GAO X Y, WEI S, WEN Z Q, et al. Citrus detection method based on improved YOLOv5 lightweight network [J]. Computer Engineering and Applications, 2023, 59 (11): 212-221. 
[20]刘思言, 王博, 高昆仑, 等. 基于R-FCN的航拍巡检 图像目标检测方法[J]. 电力系统自动化, 2019, 43 (13): 162-168. 
LIU S Y, WANG B, GAO K L, et al. Object detection method for aerial inspection image based on region-based fully convolutional network[J]. Automation of Electric Power Systems, 2019, 43(13): 162-168. 
[21]唐靓, 余明慧, 武明虎, 等. 基于改进YOLOv5的绝 缘子缺陷检测算法[J]. 华中师范大学学报(自然科 学版), 2022, 56(5): 771-780. 
TANG J, YU M H, WU M H, et al. Insulator defect de tection algorithm based on improved YOLOv5[J]. Jour nal of Central China Normal University (Natural Sci ences), 2022, 56(5): 771-780. 
[22] JIN Y, GAO H F, FAN X L, et al. Defect identification of adhesive structure based on DCGAN and YOLOv5[J]. IEEE Access, 2022, 10: 79913-79924. 
[23]秦晓辉, 黄启东, 常灯祥, 等. 基于改进YOLOv5的 露天矿山目标检测方法[J]. 湖南大学学报(自然科 学版), 2023, 50(2): 23-30. 
QIN X H, HUANG Q D, CHANG D X, et al. Object de tection method in open-pit mine based on improved YOLOv5[J]. Journal of Hunan University (Natural Sci ences), 2023, 50(2): 23-30.

更新日期/Last Update: 2024-06-14