电工技术2024,Issue(1) :142-145.DOI:10.19768/j.cnki.dgjs.2024.01.036

基于图像识别的变电站鸟巢检测技术综述

Review on Image Recognition-based Detection Technologies for Substation Bird Nests

金伟超 徐国栋 周海阔 边万聪 陈伟
电工技术2024,Issue(1) :142-145.DOI:10.19768/j.cnki.dgjs.2024.01.036

基于图像识别的变电站鸟巢检测技术综述

Review on Image Recognition-based Detection Technologies for Substation Bird Nests

金伟超 1徐国栋 1周海阔 1边万聪 1陈伟1
扫码查看

作者信息

  • 1. 国网苏州供电公司,江苏 苏州 215000
  • 折叠

摘要

随着生态环境不断改善,变电站鸟类筑巢缺陷频发,严重影响电力系统的安全运行.基于图像识别的变电站鸟巢检测技术具有巡检效率高、检测周期短的优点.对当前基于图像识别的鸟巢检测技术研究进行了综述.首先,详细介绍了鸟巢图像识别的预处理、特征提取和目标检测算法等关键技术;然后,结合苏州供电公司试验和运行结果,分析并探讨了变电站鸟巢检测技术在实际应用中的难点与痛点;最后,提出了技术和管理上的改进措施.

Abstract

With the continuous improvement of ecological environment,substations frequently encounter bird nest defects,which has become a serious threat to safe and stable operation of power system.The image recognition-based technologies for detecting substation bird nests,with the advantages of high inspection efficiency and short detection peri-od,are the focus of this review.First key technologies of bird nest image recognition such as preprocessing,feature ex-traction and target detection algorithm are introduced in detail.Then difficulties and pain points of the technologies in practical applications are analyzed and discussed based on test and operation results of Suzhou Power Supply Company.Fi-nally the improvement measures on technology and management are proposed.

关键词

变电站/鸟巢/图像识别/自动检测

Key words

substation/bird nest/image recognition/automatic detection

引用本文复制引用

出版年

2024
电工技术
重庆西南信息有限公司(原科技部西南信息中心)

电工技术

影响因子:0.177
ISSN:1002-1388
参考文献量15
段落导航相关论文