卷积神经网络的电力检修票智能识别研究  

Research on Intelligent Identification of Power Maintenance Ticket Based on Convolutional Neural Network

作　　者：叶华 赵明 王国平 赵川 YE Hua;ZHAO Ming;WANG Guo-ping;ZHAO Chuan(Yunnan Electric Power Dispatch Control Center,Kunming 650217 China)

机构地区：[1]云南电力调度控制中心,云南昆明650217

出　　处：《自动化技术与应用》2022年第1期97-101,共5页Techniques of Automation and Applications

摘　　要：以提升电力检修票智能识别准确率和识别效率为目标,提出卷积神经网络的电力检修票智能识别方法。实验结果表明,该方法能够很好完成特征点匹配;且电力检修票的智能识别准确率高、识别效率快,具备良好的使用性能。In order to further improve the accuracy and efficiency of the intelligent identification of power maintenance tickets, a convolution neural network intelligent identification method of power maintenance tickets is proposed. The experimental results show that the method can match the feature points well, and the intelligent identification accuracy of the electric overhaul ticket is high, the recognition efficiency is fast, and the performance is good.

关 键 词：卷积神经网络 电力检修票 智能识别 特征匹配 射影变换准则 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]