Research On 3D Laser Scanning Data Simplification, Surface Reconstruction Methods And Applications

Three-dimensional laser scanning technology is a new type of measurement technology that has emerged in recent years.Because of its high efficiency and accuracy,it is widely used in surveying and mapping,reverse engineering etc.The original point cloud data obtained by the 3D laser scanning technology is huge and has a large amount of redundancy,these redundant data will reduce the utilization efficiency of the point cloud,and cause inconvenience of point cloud data processing and application management.In addition,by performing surface reconstruction on the point cloud,it can provide the user with real three-dimensional scene information of the target.Therefore,studying the point cloud simplification algorithms and surface reconstruction algorithms is of great significance for improving the availability and visibility of point clouds.This paper focuses on the algorithm of point cloud simplification and surface reconstruction,and verifies the effectiveness of the proposed algorithm through programming and experimental analysis.The main contents and results of this paper are as follows:(1)The method of establishing the topological relationship of unorganized point cloud is introduced in detail.In order to solve the problem that the convergence of the classical k-means clustering method is slow and the results of multiple clustering are not uniform,this paper uses octree tree to provide initial clustering center for clustering,and then uses elkan k-means clustering method to reduce the calculation of distance.Experiments show that the proposed method has fewer iterations and higher computational efficiency than the classical k-means method,and can better achieve the clustering of scattered point clouds.(2)In order to reduce the loss of feature information in point cloud simplification,a point cloud internal feature point extraction method based on unit method extended distance is proposed.This method calculates the difference between the max distances of unit normal vectors of points and point spacing,and this value is used as a new feature detection operator to describe the degree of surface variation.Experiments show that the algorithm can accurately extract the internal features of the point cloud.(3)In order to preserve the boundary information of the point cloud in the simplification process,a point cloud boundary extraction method based on the neighboring point tangent plane projection distribution is proposed.This paper analyzes the distribution characteristics of projection points and uses multi-level judgment to identify boundary points.Experiments show that the algorithm can accurately detect the boundary points of different types of point clouds.(4)Analyzed the shortcomings of the classical point cloud simplification algorithm,and proposed a point cloud simplification algorithm based on feature preservation.Firstly,the method uses the point cloud feature extraction method to preserve the internal features and boundary features of the original point cloud,and uses the point cloud clustering algorithm proposed in this paper to initially cluster the point cloud.Then a clustering subdivision method based on variational variance of intra-surface surfaces is proposed to realize the extraction of weak feature points and the uniform simplification of non-feature points.Experiments show that the proposed method has a good simplification effect on point clouds with different geometric features,and is superior to the classical method in terms of simplification effect and accuracy.(5)Aiming at solving the problem of inconsistent normal vector direction in Poisson surface reconstruction,a method of point cloud normal vector consistency adjustment based on additional constraint information is proposed.Experiments show that this method can achieve correct adjustment of the point cloud normal vector.After adjustment,Poisson reconstruction can obtain accurate modeling results.Applying the algorithm of this paper to actual data,including point cloud simplification and surface reconstruction,verify that the algorithm has practical value.