| With the development of UAV technology,radar technology,threedimensional point cloud technology and the rapid development of China’s power industry,the inspection technology of power corridor equipped with lidar equipment is gradually replacing the previous manual inspection method because of its high efficiency,accuracy,convenience and many other advantages.It is a problem worthy of study to segment the obtained inspection point cloud,obtain the interest points such as power cable,tower and ground after segmentation,and carry out relevant calculation.Aiming at the problems of large computation and inaccurate cable extraction of traditional point cloud segmentation technology,this paper improved the existing filtering method of power inspection point cloud and proposed a new cable extraction method,which mainly includes :(1)Based on statistical method and gradient threshold,the method of combining the feature of power checking point cloud,improved statistical histogram filtering,can rapidly in large amount of data of point cloud data to extract power corridor part,more effectively reduce the subsequent computation for the work of the extraction of cables,tower,at the same time as most will filter out on the ground,It also improves the segmentation accuracy of the whole.(2)A region growth extraction method based on PCA is designed to extract power cables accurately,and the vertical plane of the segmented cable is projected.Catenary fitting is performed on the projection plane to obtain the catenary equation of the cable in three-dimensional space.(3)A method is proposed to segment tower and ground based on obtaining boundary points of tower based on minimum enveloping rectangle and fitting boundary lines with sampling consistency algorithm to calculate basic contour of tower.(4)According to the algorithm proposed in this paper,the design and implementation of power inspection point cloud segmentation system are completed,and detailed explanation is given. |
PDF Full Text Request