Research On 3D Point Cloud Data Simplification And Mesh Reconstruction Algorithm

In recent years,with the development of reverse engineering technology,3D point cloud model is widely used in virtual reality,3D map,medical image,film and television special effects,etc.How to quickly and accurately model,become a key issue of 3D reconstruction technology.Since the acquired point cloud inevitably has redundant data,in order to ensure the quality and efficiency of modeling,it is necessary to simplify the point cloud.In this paper,the point cloud data simplification and triangle mesh reconstruction in the reconstruction process are studied in depth.The specific research work is summarized as follows.Firstly,aiming at the problem that the existing point cloud simplification algorithm often loses a lot of detailed features in the simplification process,a point cloud simplification algorithm based on feature retention is proposed.The algorithm subdivides the space by fitting the flatness value of the point cloud data,introducing the normal estimation and the weighted average Gaussian kernel function to define the feature points in the feature region,and extracting the feature points;using the voxel method in the flat region for fast simplify,and then use the random simplification method to resample the point cloud,and merge it into the point cloud after extracting the feature points,and then remove the duplicate points to realize the simplification of the point cloud.Secondly,in the triangle mesh reconstruction stage,a triangle mesh reconstruction algorithm based on multi-criteria is proposed for the problem of low grid growth efficiency and low grid growth quality.Based on the region expansion method,the algorithm combines the idea of projection method.With a seed triangle as the starting condition,the projection method is used to map the adjacent points of the triangle to the two-dimensional tangent plane,and multiple criteria are designed to quickly determine the candidate.Extend the range of points,then map the two-dimensional relationship of the candidate points back to the three-dimensional space,set priorities for the candidate triangles in the space,select the optimal triangles,and continuously grow in this way to quickly reconstruct a high-quality grid model.Finally,based on the three indicators of accuracy,simplicity and speed,and the grid generation quality and algorithm operation time-consuming two indicators,the effectiveness of the reduction algorithm and the reconstruction algorithm are verified respectively.The simulation results show that the point cloud reduction algorithm based on feature retention and the multi-criteria triangle mesh reconstruction algorithm have achieved good results.