Abstract: The power distribution network′s overhead lines exhibit a slender structure on the millimeter scale, resulting in sparse point clouds during long-distance scanning, which makes it difficult to calculate the distances between power line point clouds. To address this issue, a remote monitoring method for power line safe distances using lidar point cloud fusion is proposed in this paper. First, power line point cloud data is captured and fused. Then, multi-lidar calibrated and fused point cloud features are coarsely extracted. After that, a KD-tree is constructed to calculate the distances between the point clouds, enabling precise extraction via mobile laser point fitting. That allows for the accurate extraction of power line positional information from complex point cloud data. Finally, the power line point cloud data is fitted according to the sag equation, and the vertical distances from each point on the power line to the center point of the bucket truck are obtained. Thereby, the remote automatic monitoring of the safe distance for distribution network power lines is accomplished. Experimental results demonstrate that the proposed method can accurately and completely extract the power line point cloud data, with the maximum fitting residuals for the shield wire and power line point clouds being 0.119 m and 0.224 m respectively. An alarm is triggered at a distance of 4.95 m, with a response time of 0.45 s and a monitoring coverage rate of 97.5 %, which effectively reduces operational risks.