山东电力技术2025,Vol.52Issue(5):1-8,8.DOI:10.20097/j.cnki.issn1007-9904.2025.05.001
基于改进YOLOv8的输电线路外力破坏检测算法
Transmission Line External Force Damage Detection Algorithm Based on Improved YOLOv8
摘要
Abstract
Aiming at the high requirements for both accuracy and detection speed of the target detection model in intelligent inspection of transmission lines,and the complex and variable background of transmission lines,a transmission line external force damage detection algorithm based on improved YOLOv8 is proposed.The spatial channel reconstruction convolution module is used to reduce the computational resources required for redundant feature extraction;an adaptive spatial feature fusion technique is employed to address the feature fusion of external force damage targets in complex backgrounds;A bounding box loss function based on the minimum point distance is used in the loss function to increase the detection accuracy of the algorithm.Comparison experiments with traditional algorithms and other improved target detection algorithms find that the detection accuracy and detection speed of the proposed algorithm are significantly better than other algorithms.The results show that,on the transmission line external force damage dataset,the mean average precision(mAP)at 50%confidence,the mAP at 50%to 95%confidence,and the detection speed are improved by 5.3%,9.2%,and 23.6 frames per second(FPS),respectively,which effectively enhance both detection accuracy and speed,making the proposed algorithm applicable for detecting external force damage to transmission lines.关键词
输电线路智能巡检/YOLOv8/空间通道重建卷积/渐进特征金字塔/边界框损失函数Key words
intelligent inspection of transmission lines/YOLOv8/spatial and channel reconstruction convolution/asymptotic feature pyramid network/MPDIoU分类
计算机与自动化引用本文复制引用
赵文清,蔡建颖,马玉梅,翟永杰..基于改进YOLOv8的输电线路外力破坏检测算法[J].山东电力技术,2025,52(5):1-8,8.基金项目
国家自然科学基金项目(62371188).National Natural Science Foundation of China(62371188). (62371188)