Abstract
In order to further improve the anomaly detection level of transmission lines, a transmission line three-dimensional modeling method based on 3D laser scanning point cloud data was proposed. The AP clustering algorithm based on density features was used to segment and extract power lines, so as to obtain high-precision power line point cloud data. Then, the 3D model was transformed into a 2D plane model. Finally, the above modeling method was applied to the tree barrier detection of 2D kd-tree. The results showed that AP clustering based on density feature shows good clustering effect when segmenting and extracting power lines. Compared with other clustering algorithms, the proposed clustering algorithm shows good clustering effect in terms of classification rate, adjusted Rand index, silhouette coefficient and running time, and the accuracy reaches 99.5678%. Combined with the application of tree barrier detection of kd-tree, the proposed modeling method has higher detection accuracy and efficiency in tree barrier detection, and it only takes 7.68s to complete the detection. Therefore, the proposed modeling method shows good performance and application effect, which provides a new reference for the application of 3D modeling technology in power engineering.