基于移动边缘计算的电网输电线路巡检图像识别研究

Research on Image Identification of Power Transmission Line Inspection Based on Mobile Edge Computing

针对传统的电网输电线路巡检图像识别方法中资源和时间消耗较大的问题,提出基于移动边缘计算的电网输电线路巡检图像识别研究。使用移动边缘计算技术,设计边缘数据处理层,在线处理部分障碍严重的巡检图像,通过图像预处理技术处理剩余巡检图像,利用图像投影提取输电线路巡检图像特征,将获得的特征抽象到二维平面中,实现电网输电线路巡检图像识别。测试结果表明,与传统的识别方法相比,设计的基于移动边缘计算的电网输电线路巡检图像识别方法大幅度缩减了资源和时间的消耗,该方法更适合应用在电网输电线路巡检图像识别中。

Aiming at the problem of large resource and time consumption in the traditional power transmission line inspection image recognition method,a mobile edge computing based power transmission line inspection image recognition research is proposed.Using mobile edge computing technology,design the edge data processing layer,process some of the inspection images with severe obstacles online,process the remaining inspection images through image pre-processing technology,use image projection to extract the features of the transmission line inspection images,and abstract the obtained features in the two-dimensional plane,the image recognition of the grid transmission line inspection is realized.The test results show that compared with the traditional identification method,the proposed image identification method based on moving edge calculation for power grid transmission line patrol greatly reduces the consumption of resources and time,and this method is more suitable for power grid transmission line patrol image identification.