自动化与仪表2024,Vol.39Issue(12) :100-104.DOI:10.19557/j.cnki.1001-9944.2024.12.022

基于视觉识别技术的电能计量装置状态自动化检测系统

Automatic Detection System for the Status of Electric Energy Metering Device Based on Visual Recognition Technology

韩瑞 李文娟 杨生婧 索吉鑫
自动化与仪表2024,Vol.39Issue(12) :100-104.DOI:10.19557/j.cnki.1001-9944.2024.12.022

基于视觉识别技术的电能计量装置状态自动化检测系统

Automatic Detection System for the Status of Electric Energy Metering Device Based on Visual Recognition Technology

韩瑞 1李文娟 1杨生婧 1索吉鑫1
扫码查看

作者信息

  • 1. 国网青海省电力公司信息通信公司,西宁 810000
  • 折叠

摘要

针对目前电能计量装置状态检测存在检测难度较大、成本高以及安全性低等问题,提出基于视觉识别技术的电能计量装置状态自动化检测系统.系统以视觉识别技术为依据,创新地提出混合图像校正方式,通过OCR模型进行数据识别处理;通过迁移学习对数据集进行训练,以残差值范围进行电能计量装置状态异常与否的判断.设计了针对视觉识别效果以及电能计量装置状态自动化检测效果的实验,实验结果表明系统能够准确地识别电能计量装置图像,数据处理准确高效,电能计量装置状态检测可靠灵活,准确性高.

Abstract

In order to improve the safety,flexibility and reliability of automatic state detection of electric energy me-tering device,an automatic detection system of electric energy metering device state based on visual recognition tech-nology is proposed.Based on the visual recognition technology,the system innovatively proposes a hybrid image cor-rection method,and performs data recognition and processing through the OCR model.The dataset was trained by transfer learning,and the residual value range was used to judge whether the state of the electric energy metering device was abnormal.An experiment was designed for the visual recognition effect and the automatic detection effect of the state of the electric energy metering device.The experimental results show that the automatic detection system of electric energy metering device status based on visual recognition technology can more accurately identify the im-age of electric energy metering device,the data processing is accurate and efficient,and the state detection of electric energy metering device is reliable,flexible and accurate.

关键词

OCR模型/视觉识别技术/迁移学习/累加判断/电能计量装置

Key words

OCR model/visual identity technology/transfer learning/accumulated judgment/energy metering device

引用本文复制引用

出版年

2024
自动化与仪表
天津市工业自动化仪表研究所 天津市自动化学会

自动化与仪表

CSTPCD
影响因子:0.548
ISSN:1001-9944
段落导航相关论文