Obstacle avoidance during aerial inspection of power lines

Abstract

Gives an overview of a research programme that aims to apply machine vision for guiding a small, unmanned helicopter during inspection of overhead electrical distribution lines. Briefly discusses the background and requirements for aerial inspection of power lines and states the advantages of using a remotely piloted vehicle (RPV). Identifies the main obstacle to the use of RPVs in this application as the ’‘see and avoid” principle, which arises from regulatory requirements; it is the underlying motivation for the work described here. Machine vision and automated path planning offer a potential solution. Gives a brief tutorial of the principles involved and describes research in image processing and rapid path planning aimed at detecting and avoiding obstacles in the airspace of a small RPV. Also presents experimental results from a laboratory test rig which was constructed to assess the methods.