In this work, an automatic power line extraction method based on airborne LiDAR point cloud data is proposed. First, spatial partitioning of LiDAR data was performed. Second, according to the horizontal distribution characteristics of power lines in three-dimensional space, an improved Euclidean clustering algorithm was used to realize rough extraction of the power lines. Third, using the connection between a power line and a power tower, the spatial coordinate position at the top of the power tower was estimated. Then, the improved Euclidean clustering algorithm was used to realize single power line extraction, and the model was used to combine a straight line and parabola to obtain the centerline equation of a single power line and its radius. Finally, a power line adapter was developed at the insulator according to the power line equation and radius, and the complete point cloud of a single power line was obtained. Experiment results show that compared with the classification effect of support vector machines combined with the geometric feature method, the proposed method can extract complete power lines automatically, quickly, and accurately from power line inspection data, which has application value in power patrol.