Power Lines Extraction and Distance Measurement from Binocular Aerial Images for Power Lines Inspection Using UAV

With the development of autonomous perception and obstacle avoidance technologies for power lines inspection using unmanned aerial vehicles(UAV), the methods of power lines extraction and distance measurement from binocular aerial images become the spotlights gradually. In this paper, the Ratio detection operator and fast Hough transform are adopted to extract the pixels of power lines and detect straight lines respectively, a pair of straight lines are selected by demand-based strategy further. Then, aim at the distance measurement of power lines, a method of constructing homonymy points based on the linear hypothesis of power line and epipolar geometry constraint is proposed. The three dimensional(3D) relative coordinates of two spatial points on the power line are calculated by the image coordinates of two pairs of homonymy points and camera intrinsic parameters. Last, the shortest distance between UAV(left camera) and power line determined by two spatial points is calculated. The experiment results show that power lines are extracted accurately and the process of distance measurement have good real-time performance.