摘要
文中针对目前无人机巡检过程中产生的海量图像和视频数据在人工处理时工作效率低下、工作量巨大等问题,采用J2EE技术体系,按照多层次软件开发体系对输电线路无人机智能管理系统进行了研究。该系统采用了基于Faster-Rcnn的深度学习算法,利用深度卷积神经网络算法对数据进行预处理和分类。同时能够完成设备缺陷的标注,并将识别结果进行反馈。此外,系统集成了无人机巡检工作的整个作业流程,实现了对无人机巡检作业的智能化调度和全流程监控。该系统的研究对于实现输电线路无人机巡检的高效工作具有重要意义。
In this paper,aiming at the problems of inefficiency and huge workload in manual processing of massive images and video data produced in the process of UAV patrol inspection,the intelligent management system of UAV on transmission line is studied by using J2 EE technology system and multi-level software development system. In this system,the depth learning algorithm based on Faster-Rcnn is adopted,and the data is pre-processed and classified by the depth convolution neural network algorithm.At the same time,the equipment defect is marked and the recognition result is feedback. At the same time,the system integrates the whole work flow of UAV patrol work,and realizes the intelligent scheduling of UAV patrol work and the whole process monitoring. The research of the system is of great significance for the efficient work of transmission line UAV inspection.
作者
郑仟
李宁
ZHENG Qian;LI Ning(Maintenance Company,State Grid Ningxia Electric Power Co.,Ltd.,Yinchuan 750011,China)
出处
《电子设计工程》
2019年第9期74-77,82,共5页
Electronic Design Engineering
关键词
无人机巡检
深度学习
卷积神经网络
缺陷识别
信息管理
UAV inspection
deep learning
convolution neural network
defect recognition
information management