To address the issues of low automation and segmentation errors caused by parameters in power line point cloud extraction, this paper proposes a power line extraction method based on an improved spatial density clustering algorithm, combined with the distribution characteristics of airborne LiDAR point cloud data. Firstly, the proposed method completed the rough extraction of power line point cloud through the improved elevation filtering algorithm. Then, the optimal parameters of spatial density clustering were obtained by the distance-density method and the mathematical expectation calculation method, avoiding the complicated manual parameter adjustment process. Experimental results show that compared with the spatial density clustering algorithm, the proposed algorithm has significantly improved efficiency, reduced the power line extraction time by about 60%, and realized the automatic and efficient extraction of single power line point cloud.