SIFT algorithm based real-time power system inspection by using unmanned aerial vehicles

Abstract

This research presented a novel method for transmission and distribution lines inspection. Power lines inspection can guarantee the performance of power system stability and increase the lifetime of power lines. An advanced power lines inspection can prevent massive power outages due to line failures, which is still one of the major challenges in modern power system. Traditiional method is using human squads to perform inspections. Due to the remoted locations of power lines, and various geographical conditions, it is not hte most efficient solution by dispatching human squads. Another popular method is using helicopters to inspect power lines, which is costly and there are limitations due to weather conditions. In this research, a combination of Scale Ivariant Feature Transform algorithm and Unmanned Aerial Vehicles (UAVs) power lines real-time inspection system is presented. SIFT algorithm is widely used in object recognition, and it can help USAVs to reduce the vibration problems due to wind or other weather issues, while videoing power lines. The advantage of SIFT method of being invariant to image translation, scaling and rotation has been fully implemented in the proposed work. The video is processing to ground station and the final inspection analysis is performed frame by frame to locate or to recognize potentioal unhealty power lines.