Research On Boundary Detection And Hole Repairing Based On 3D Laser Scanning Point Cloud

3D laser scanning technology is a technological innovation of Surveying and mapping science. It has become a hot spot in many fields with its unique advantages. Due to the influence of factors such as the object shelter or complex object surface in the process of scanning, measurement data inevitably produce holes, which must be to affect the quality of reconstruction after the digital model. Studying on point cloud boundary detection and hole repairing can ensure the integrity of the point cloud data, that is advantageous to subsequent data processing and model reconstruction. This article focus on the corresponding algorithm in the hole repair process, through the programming implementation and experimental test, the improvement of the related algorithms are completed. The major work and achievements of this paper are as follows:(1) In view of the irregular distribution of scattered point cloud data, this paper proposes an improved k-nearest neighbors searching algorithm based on dynamic grid decomposition, which can establish point cloud topological relations. Experiments show that this algorithm can not only search the k neighboring points of the object point fast and exactly, but also has a strong applicability.(2) In order to detect the boundary of the point cloud data, a method based on the lifting method is proposed. This method gives target point and adjacent points to the same weight, and regards the target as a fulcrum for suspension. According to the spatial distribution of each point in the physical performance distribution whether discriminant target is the boundary point. The experimental results show that this method can accurately detect the boundary points.(3) In order to distinguish the inner and outer boundaries of the point cloud data, this paper researches and analyses of the relationship between inside and outside of the border and other data points, based on the centroid of the inside and outside of the border and other data points with respect to the location of the boundary of different, distinguish the inner and outer boundary. Experiments show that the method can accurately determine the inner and outer boundary.(4) In order to repair the hole area of the point cloud data, the correlation algorithm is studied, and the hole area is repaired by the improved method based on the radial basis function. It projects the hole boundary points to the characteristic plane, and carries on the characteristics of surface interpolation based on scan line, then uses the radial basis function for curved surface fitting. Finally, it maps the new points on the feature plane to the fitting surface to realize the hole repairing. Experiments showed that the method in dealing with different curvature change region has good repairing result, and has strong applicability.(5) Through applying this algorithm to Guishan Tomb and the gas tank in fengfeng mining area, repairing the hole of the point cloud data, to verify this algorithm has certain practicality.