基于深度学习的电力设备铭牌文本识别

电力设备铭牌包含丰富的设备信息，通过图文识别技术获取设备铭牌信息，可更加高效快捷地完成电力设备的信息统计、台帐校核等工作，也有利于提高电力系统的设备管理水平。针对电力设备铭牌与普通图像文本识别差异较大的特殊应用场景，文中提出一种基于深度学习的电力设备铭牌信息自动识别算法。该算法由铭牌检测、文本检测、文本识别三部分组成。通过改进损失函数设计、增加文本识别结果纠正、人工合成文本图像等方式，使得铭牌检测模型在测试集上的平均精度均值达到92.2%，文本检测模型在测试集上的F₁值达到91.2%，文本识别模型的字符识别准确率达到94.0%，文本行识别准确率达到82.3%。

Text recognition of power equipment nameplates based on deep learning

There is abundant information in power equipment nameplates. Extracting information from nameplates through image and text recognition technology enables the high effective and quick works on statics and account check of power equipment,which is also beneficial to improve the equipment management level of power system. Considering the great difference of text recognition between power equipment nameplates and ordinary images,an algorithm of automatic recognition of power equipment nameplates based on deep learning is proposed in this paper. This algorithm consists of three parts,namely nameplate detection,text detection and text recognition. By improving the design of loss functions,adding the correction of text recognition,and synthesizing text images,the mean average precision of the nameplate detection model on the test set reaches 92.2%,the F₁ of the text detection model on the test set reaches 91.2%,and the character recognition accuracy rate of the text recognition model reaches 94.0%,the text line recognition accuracy rate of the text recognition model reaches 82.3%.