Research On The Key Technologies Of 3D Modeling Based On Terrestrial Laser Scanning

The ground laser radar(LIDAR)as a kind of advanced precision 3d measurement tools,the measured data of point cloud,simple in structure and details of surveying and mapping precision,can be accurately reconstruct the object surface.The amount of points in these point cloud rising rapidly when laser scanners are widely using.The registration and surface reconstruction of the “Mass Point Cloud” as the key technologies of 3D modeling,selection and optimization mode ling directly affect the efficiency of modeling is the current research hot spot.In this paper,registration algorithms of Point cloud data,which are obtained with the laser scanner,are investigated.Here in,an ICP algorithm,which uses k-d tree and angle between the direction vectors of two point cloud,is presented to improve the classical ICP algorithm.Experimental results are presented to prove that the proposed algorithm can improve the speed of registration and reduce the error rate,moreover it avoids local minimum.The main content of this study is organized as follows,(1)Firstly,algorithms to process the scanning point cloud data are investigated in detail.With the property of large in scale and non-direct relation between different points,the process of point data and the topology depends on the neighborhood.A great number of operations for query and update are involved to establish the topological relationship.Storage the data points in the k-d tree is a good idea,as this paper does.(2)Secondly,according the defect of ICP algorithm,the paper proposes an improved ICP algorithm for point cloud registration.First of all,the initial registration in this paper uses principal component analysis method to obtain a better initial position of point cloud.Then two point clouds are registered by k-d tree search and angle between the direction vectors of two point cloud based on traditional ICP algorithm.The comparison and analysis is made with the classical ICP algorithm.The accuracy of the improved process of algorithm will not be affected to improve the efficiency.The algorithm is verified in the applications.(3)Crust algorithm is further proposed to realize to Delaunay triangulation and modeling of point cloud data.Also the algorithm can easily to programming and save the feature information of the target object with high efficiency.The algorithm is verified in the applications.