基于语义分割技术的输电线路识别
Identification of Transmission Lines Based on Semantic Segmentation Technology
管诗衡 1皮梦婷 1胡为民 1余杰 1朱佳1
作者信息
- 1. 国家电网宜昌供电公司,湖北宜昌 443002
- 折叠
摘要
输电线路的安全关乎国民生产生活的正常用电,因此,输电线路的巡检显得尤为重要,主要是对电线进行状态检测和故障诊断.基于语义分割算法对输电线路进行分割是输电线路巡检的一种有效手段.与传统的机器学习算法相比,语义分割方法能更有效地实现输电线路的识别和缺陷检测.文章基于语义分割技术,使用不同语义分割算法对输电线路数据集进行分割,并分析总结了这些算法在输电线路数据集上存在的问题.
Abstract
The safety of transmission lines is related to the normal use of electricity in national production and life,so the patrol inspection of transmission lines is particularly important. It can conduct status detection and fault diagnosis on the lines. Transmission line segmentation based on a mantic segmentation algorithm is an effective means of transmission line inspection. Compared with traditional machine learning algorithms segmentation method can more effectively realize transmission line identification and defect detection. Based on semantic segmentation technology,this paper uses different semantic segmentation algorithms to segment transmission line data sets and analyzes and summarizes the problems of these algorithms in transmission line data sets.
关键词
输电线路/语义分割/机器学习/图像处理技术Key words
transmission line/semantic segmentation/machine learning/image processing technology引用本文复制引用
出版年
2024
NETL
NSTL
万方数据