Image identification method and system for identifying abnormal label of electric energy meter

• Main Authors: ZHANG LIJUAN, ZHANG CHUN, LAI GUOSHU, SU DONGSHENG, ZHANG QIONGHAI, LIN SUCUN, HUANG JIAOJIAO, LI JIANXIN, WANG JIAO, HONG QIAOWEN, ZHOU HOUYUAN, ZHAN MENGTAO
• Format: Patent
• Language: Chinese; English
• Published: 09.11.2021
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; HANDLING RECORD CARRIERS; PHYSICS; PRESENTATION OF DATA; RECOGNITION OF DATA; RECORD CARRIERS

The invention relates to an image identification method for identifying an abnormal label of an electric energy meter, and the method comprises the following steps: S1, obtaining a label image, and carrying out the preprocessing of the label image, and obtaining a sample data set; S2, constructing an abnormal label identification model based on the residual network, and training based on the sample data set; and S3, inputting a to-be-identified tag image into the trained abnormal tag identification model to obtain an abnormal prediction result, namely an abnormal identification result corresponding to the to-be-identified tag image. According to the invention, abnormal label identification efficiency and accuracy can be considered in operation environments such as an automatic production line, so that an abnormal label identification effect is improved. 本发明涉及一种用于电能表异常标签识别的图像识别方法，包括以下步骤:步骤S1:获取标签图像，并预处理，得到样本数据集；步骤S2:基于残差网络，构建异常标签识别模型，并基于样本数据集训练；步骤S3：将待识别标签图像输入经过训练后的异常标签识别模型中，得到异常预测结果，即为对应待识别标签图像的异常识别结果。本发明能够在自