The Deep Learning Based Power Line Defect Detection System Built on Data Collected by the Cablewalker Drone

Traditional technologies used for servicing and inspecting power line condition are expensive, time consuming and put people's lives at risk. Employing UAVs equipped with payloads for power line inspection combined with machine vision and machine learning techniques to analyze the data collected can offer great improvements in terms of defect detection efficiency, cost and safety. The article talks about the neural network created to detect and classify power line poles on images or videos. The neural network in question is the first part of the larger end-to-end power line defect detection software being developed.