UAV Power Line Detection and Tracking using a Color Transformation

Power line inspection assists with fault finding and regular maintenance strategy, which are important functions from an operational perspective. Modern technologies that have been researched over the last decade have shown the potential for this process to become automated with the use of unmanned aerial vehicles, which can reach areas that are just too difficult to reach by common conversions. For an unmanned aerial vehicle to successfully monitor and inspect a power line, it needs to be equipped with a proper power line detection and tracking algorithm. There have been various techniques proposed over the years as this research area is continuing to grow and mature. A practical power line detection and tracking using color transform, as the main algorithm, is presented here. The work is done using active image processing, aided by python OpenCV. On each frame of the image, the necessary camera calibrations are performed, and thereafter a perspective transform is carried out to obtain the bird's eye view. A color transform is then used to isolate clutter from the lines, which enables the detection of actual lines, together with the calculation of the necessary distances from the lines. This implementation and preliminary findings of this experimental method are given here, which indicate that emulated power lines can be successfully tracked using the presented approach.