高电压技术2024，Vol.50Issue(5)：1855-1864,10.DOI:10.13336/j.1003-6520.hve.20220387
基于改进YOLOv5-LITE轻量级的配电组件缺陷识别
Defect Identification of Distribution Components Based on Improved YOLOv5-LITE Lightweight
摘要
Abstract
In order to accurately and quickly locate and identify the defects of distribution components,a lightweight defect identification method of distribution components based on improved YOLOv5-LITE is proposed.To make the model easy to deploy to mobile device terminals,this method uses Shufflenetv2 as the backbone network to extract fea-tures,constructs YOLOv5-LITE lightweight neural network model,and removes 1024 convolution and 5×5 Pooling of Shufflenetv2,which is replaced by global average pooling operation to reduce the amount of network parameters and im-prove the speed of model detection.By introducing the 152×152 feature layer,which is conducive to the detection of fine-grained objects,the defect detection of large-,medium-and small-scales is realized.Using deep separable convolu-tion instead of downsampling in PANet architecture makes the network more lightweight.The experimental results show that this method can be adopted to identify three defects:cable separation gasket,cable and insulator falling off and acy-clic insulator.The detection accuracy is 92%,95%,and 95%,respectively.The amount of network parameters is about 1/4 of YOLOv5,and the detection speed is 2 ms/piece.The proposed method has the characteristics of real-time,high accu-racy and light weight.关键词
目标检测/YOLOv5/ShuffleNetV2/轻量化/配电线路/缺陷识别Key words
target detection/YOLOv5/ShuffleNetV2/lightweight/distribution line/defect identification引用本文复制引用
颜宏文,万俊杰,潘志敏,章健军,马瑞..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):1855-1864,10.基金项目
国家自然科学基金(51977012) （51977012）
国网湖南电力科技项目(5216A32100AF).Project supported by National Natural Science Foundation of China(51977012),Science and Technology Project of State Grid Hunan Electric Power(5216A32100AF). （5216A32100AF）
