电子器件2025，Vol.48Issue(2)：426-431,6.DOI:10.3969/j.issn.1005-9490.2025.02.029
基于多尺度序列视觉特征融合模型的输电线路防外破红外感应预警技术
Infrared Induction Early Warning Technology for Transmission Line Breaking Prevention Based on Multi-Scale Sequence Visual Feature Fusion Model
摘要
Abstract
Targeting at the randomness and high risk of transmission line external damage,an infrared induction early warning technology for transmission line external damage prevention based on multi-scale sequence visual feature fusion model is proposed.The main sta-tion,base station,infrared camera and other equipments are used to form an infrared induction system to collect the image of the trans-mission line.Atmospheric scattering model is used to enhance image contrast and improve image quality.Gaussian pyramid model is used to extract the color and brightness features of different scale sequences in the image.The attention mechanism is introduced to cal-culate the maximum value of saliency.Different scale features are fused according to the size of the target area,and linear transformation is made to detect and identify whether the target has external force damage points.If there are safety hazards,the early warning system will be triggered.The experimental results show that the proposed method not only improves the quality of infrared image,but also re-duces the error warning events,which is conducive to improving the safety of transmission lines.关键词
多尺度序列/视觉特征融合/输电线路/防外破/红外感应Key words
multiscale sequence/visual feature fusion/transmission line/external damage prevention/infrared induction分类
信息技术与安全科学引用本文复制引用
杜迎春,王少杰,侯婷婷,马陆阳..基于多尺度序列视觉特征融合模型的输电线路防外破红外感应预警技术[J].电子器件,2025,48(2):426-431,6.基金项目
宁夏电力有限公司2022年科技项目(5229WZ220004) （5229WZ220004）
