湖南电力 ›› 2024, Vol. 44 ›› Issue (1): 53-58.doi: 10.3969/j.issn.1008-0198.2024.01.008

• 新技术及应用 • 上一篇    下一篇

基于深度学习的输电线路巡检无人机图像分类方法

杨洪朝1, 席燕辉2, 项胜2, 徐志康2   

  1. 1.长沙民政职业技术学院软件学院,湖南 长沙 410004;
    2.长沙理工大学电气与信息工程学院,湖南 长沙 410114
  • 收稿日期:2023-09-15 修回日期:2023-10-30 出版日期:2024-02-25 发布日期:2024-03-11
  • 通信作者: 杨洪朝(1977)男,汉族,博士,研究方向为智能控制、工业信息化管理、复杂信息分析与计算。
  • 基金资助:
    国家自然科学基金项目(7217010719,72171026);湖南省自然科学基金项目(2022JJ60103);中国高校产学研创新基金—新一代信息技术创新项目(2021ITA10023);湖南省科普专题项目(2022ZK4045)

Image Classification Method Based on Deep Learning for Inspection Unmanned Aerial Vehiclesin Transmission Line

YANG Hongzhao1, XI Yanhui2, XIANG Sheng2, XU Zhikang2   

  1. 1. School of Software, Changsha Social Work College, Changsha 410004, China;
    2. School of Electrical and Information Engineering, Changsha University of Science and Technology, Changsha 410114, China
  • Received:2023-09-15 Revised:2023-10-30 Online:2024-02-25 Published:2024-03-11

摘要: 由于点云数据的不规则性,对点云数据特征值的分类精度很难达到输电线路巡检要求。为此,提出基于图信号处理的激光雷达点云分类方法。该方法在处理不规则的输电线路图像信号时,可直接从原始点数据中生成特征值,从而提高分类精度。为了验证所提出方法的有效性,采取实物仿真的方式对算法进行验证,结果表明图像中的分类精度可达到90%。

关键词: 输电线路巡查, 点云数据处理, 视觉图像处理, 无人机巡检

Abstract: Due to the irregularity of point cloud data, the classification accuracy of point cloud data feature values is difficult to meet the requirements of transmission line inspection. Therefore,a lidar point cloud classification method based on graph signal processing is proposed. When processing irregular transmission line image signals, the method can directly generate feature values from the original point data, thereby improving classification accuracy. In order to validate the proposed method, the algorithm is validated through physical simulation, and the results show that the classification accuracy in the image could reach 90%.

Key words: transmission line inspection, point cloud data processing, visual image processing, unmaned inspection

中图分类号: