Power Line Detection from Aerial Vehicles using Hough Transform

Automatic inspection and maintenance of remotely located power stations and transmission lines through unmanned aerial vehicles captured infrared and/or visible band image analysis is an important task. In this paper, an effective method based on fine scale edge map and Hough Transform is presented for automatic detection of power lines from infrared and visible light images. Publicly available real infrared and visible light images were used to evaluate the method and reasonably high performance is obtained (94.00% for infrared and 83.40% for visible light images). Even when the power lines are not very distinct in the images, the method detects them to some extent. The number of false negatives is comparatively less than the number of false positives, which is undoubtedly better.