Research On 3D Reconstruction Method Of Scattered Cloud Data

3D reconstruction technology is a research field that intersects with many disciplines such as computer graphics,industrial manufacturing,scientific computing visualization,medical bionics,etc.Its main research content is how to restore the geometric information of a 3D model under a certain data structure into 3D graphics and images.It can be identified by computer storage and finally it can realize quantitative analysis,multi-angle display and simple processing of the rebuilt model.Due to its wide range of disciplines and strong practicality,this area has made great progress.How to better recognize the three-dimensional model has always been a research hotspot for domestic and foreign scholars.The 3D model reconstruction technology can reconstruct a variety of data structure models,including point cloud data,images,and 3D contour lines.With the development of 3D sensing technology in recent years,point cloud data has become a 3D object in many applications and around it.The important carrier of the environment,recording the position information of the point in space,can be rapidly scanned and acquired by instruments such as a three-dimensional coordinate scanner.Because it is not doped with other additional information during reconstruction,it is a more general form of data.The traditional imagebased 3D reconstruction method loses the position information in the threedimensional space.It has unavoidable deficiencies in accuracy,reconstruction speed,and algorithm applicability.It has been difficult to satisfy people's highprecision,realistic three-dimensional models.Modeling and drawing requirements.With the continuous development of science and technology,the three-dimensional reconstruction technology based on point clouds has been gradually reconstructed as it can directly reconstruct a highly realistic threedimensional model through discrete points on the surface of the object,and has become the mainstream of the three-dimensional model reconstruction field.Based on the relevant literature of the researcher,the paper explores the reconstruction method of 3D model based on point cloud data,and proposes a new 3D reconstruction method of point cloud.This method combines the regional growth method and the mapping method in the reconstruction of point cloud three-dimensional surface reconstruction.First,the regional growth method is used to slice the point cloud data,and the generated triangle is subjected to an angle constraint,which enhances the generated triangular network.The quality of the grid.After that,the mapping method is adopted for the data after the fragmentation,so that a corresponding Delaunay triangulation network can be quickly generated,and the connection rules are added,and the rear boundary areas of different fragmentation are connected.Finally,the reconstructed surface is optimized to ensure that voids and overlapping patches do not appear.The main contributions of the method in this article are as follows:1.Combining with the regional growth method and the mapping method in the reconstruction of point cloud three-dimensional surfaces,2.Use the region growing method to quickly segment the point cloud data,and determine the curvature according to the angle between the spatial normal vectors of adjacent triangle mesh patches,thereby avoiding the calculation of the normal vector of the point cloud data.3.Two-dimensional mapping of fragmented point cloud data using simple mapping method greatly reduces the time required for Delaunay triangulation and improves overall efficiency.The experimental results in this paper show that the algorithm can construct a triangular mesh with better quality and effectively avoid distortion.By comparing the results of this algorithm with other algorithms,we found that this algorithm can quickly construct triangular facets that are close to regularization.Generate a comprehensive reconstruction of the quality of the results.