| Current inspections of power corridor systems are mainly performed by ground personnel or helicopters equipped with optical cameras.This situation not only consumes a lot of manpower and material resources,but also leads to inaccurate power line information.LiDAR can acquire three-dimensional spatial information of ground scenes,and at the same time,with the reduction of flight costs,the development space for using LiDAR for power corridor inspections has been expanded.This paper first uses the on-board LiDAR system to obtain 3D point cloud data,and extracts the classification features of LiDAR point cloud based on the characteristics of LiDAR data.Then use the random forest to classify the extracted features,analyze and evaluate the power consumption security status of the gallery.Based on the current application of airborne LiDAR technology in the power line inspection,the research work in this paper mainly includes the following parts:point cloud data feature extraction,point cloud feature classification based on random forest,point cloud processing,and power line extraction,power line fitting,and error evaluation.The details are as follows:First,we classify the point cloud into several categories based on the characteristics of the point cloud.Second,according to the point cloud we obtained,the power line point cloud is pre-processed,the power line candidate points are selected,and then the power line is extracted and fitted.Finally,we performed a fitting and error analysis on a single power line.Using this idea,we also performed a fitting and error analysis on a two-split power line.The accuracy is basically the same as that of a single power line.For multiple power lines,we use the same method,analyze the existing problems and propose corresponding improvements using k-means,and take the corresponding experiments with six-split power lines as an example,indicating that the method can be applied to multi-split power lines classification and fitting. |
PDF Full Text Request