1College of Automation Engineering, Shanghai University of Electric Power, Shanghai 200090, China
2Department of Computing Science, Curtin University, Perth 6102, Australia
Due to the limitation of measuring equipment or the shape characteristics of the model itself, the point cloud data in the triangulated grid generated by three-dimensional (3D) laser scanning point cloud often contains holes, which brings obstacles to the subsequent 3D reconstruction. Aimed at handling the problem of hole repairing, a hole surface repairing algorithm based on laser triangulation point cloud is proposed in this paper. First, the boundary of the triangular patch is determined by traversing the triangular mesh for the closed hole, and the hole is detected. Second, a new triangular patch is generated quickly at the hole polygon based on the minimum angle method to form the initial mesh. Third, the least square network and radial basis function implicit surface are combined. Besides, the curvature of the surface is minimized by using the minimum second derivative. The surface newly generated is consistent with the trend of the original mesh curvature. Finally, the hole repairing of laser point cloud is realized. Experimental results show that, compared with other point cloud repair methods, this method not only reduces the repair error, but also is suitable for hole repairing of various triangular mesh models.