Abstract:Automatic positioning of transmission towers based on laser point cloud is the foundation of UAV and helicopter inspection application. But now, tower positioning from laser point cloud mainly depends on manual processing of low efficiency. Aiming at solving automatic tower positioning problem with laser point cloud, this paper proposed a multi-dimensional feature analysis method based on 2D grid according to characteristics of transmission tower of point cloud data, such as high density, large slope and height difference. Firstly, the laser point cloud is denoised with statistical outlier removal filter and regularized with voxel grid filter, thus next feature map is computed based on processed point cloud. Then combined with superpositions of overhead line corridor area, the transmission towers are iteratively positioned according to density, height difference and slope features and locations of towers are automatically positioned from UAV laser point cloud. The proposed method is validated using transmission power corridor datasets obtained with large-scale unmanned helicopter of Guangdong Grid Co., and result shows that the proposed algorithm has high degree of automation and processing efficiency. It could promote intelligent performance of UAV and helicopter power line corridor inspection.

Research on Automatic Positioning Algorithm of Power Transmission Towers Based on UAV LiDAR