基于AI图像识别的输电线路缺陷自动识别技术研究

Research on Automatic Identification Technology of Transmission Line Defects Based on AI Image Recognition

针对传统识别技术在对输电线路缺陷问题识别时存在识别用时长,进一步造成识别效率低的问题,开展基于AI图像识别的输电线路缺陷自动识别技术研究。通过输电线路缺陷识别部件标注、基于AI图像识别的输电线路缺陷识别数据库建立、自动识别输电线路视频图像缺陷目标,提出一种全新的识别技术。通过实地场景验证证明,该识别技术与传统识别技术相比,有效缩短对输电线路缺陷问题的识别时间,并且识别准确率可达100%,为供电企业输电提供安全保障。

In view of the long time required for identification of transmission line defects in traditional identification technology,which further causes the problem of low identification efficiency,research on automatic identification of transmission line defects based on AI image recognition is carried out.Through the labeling of transmission line defect recognition components,the establishment of a transmission line defect recognition database based on AI image recognition,and automatic identification of transmission line video image defect targets,a new recognition technology is proposed.Field scene verification proves that compared with traditional identification technology,the identification technology effectively shortens the identification time of transmission line defects,and the identification accuracy rate can reach 100%,which provides safety guarantee for power transmission enterprises.