基于深度学习的风场关口计量表图像识别系统研究  

Research on image recognition system of wind pass meter based on deep learning

作　　者：肖文君 朱群强 张宇舟 昃萌 XIAO Wenjun;ZHU Qunqiang;ZHANG Yuzhou;ZE Meng(Hunan Vocational College of Labor and Personnel,Changsha 410100,China;Hunan Guozhiyun Co.,Ltd.,Changsha 410100,China;Guangzhou Mingde Electric Power Technology Co.,Ltd.,Guangzhou 510475,China)

机构地区：[1]湖南劳动人事职业学院,湖南长沙410100 [2]湖南省国智云有限公司,湖南长沙410100 [3]广州明德电力技术有限公司,广东广州510475

出　　处：《无线互联科技》2024年第18期87-90,共4页Wireless Internet Science and Technology

基　　金：2023年湖南省教育厅科学研究项目一般项目,项目名称:基于深度学习的风场关口计量表图像识别系统的研究,项目编号:23C0695。

摘　　要：文章对基于深度学习的风场关口计量表图像识别系统进行了研究,主要是识别关口计量表上参数,无须对仪器进行改造。文章以先进的云计算和数据分析为辅助并结合关口计量表采集数据,开发一套简单方便的微信小程序轻应用系统。该系统识别目标准确率高,图片处理时间快,且不直接与计量表计对接,具有完备的电力生产系统网络安全性,且通过光学图片采集信息,可屏蔽生产场所产生的电磁干扰,对发电企业起到挖掘数据能效、开发运维潜力、提升管理质量的效果。This research mainly focuses on the image recognition system of wind field pass meter based on deep learning,which is mainly to identify parameters on the pass meter without the need to transform the instrument.Assisted by advanced cloud computing and data analysis,combined with the pass meter to collect data,a set of simple and convenient wechat mini program light application system is developed,with high target identification accuracy and fast image processing time.And does not directly docking with the metering meter,with a complete power production system network security,and through the optical picture collection of information,can shield the electromagnetic interference in the production site,the power generation enterprises to tap data energy efficiency,development of operation and maintenance potential,improve the quality of management effect.

关 键 词：深度学习 关口计量表 图像识别系统 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]