Research On Geometry Processing Technology For Triangular Mesh Segmentation

Triangle mesh segmentation is an important field in 3D model processing.For the initial collection of 3d mesh model,it lacks certain semantic information and structural features.Mesh segmentation can divide the mesh into several parts or patches with physical significance by geometric signal,which helps us to further mesh compression,texture mapping,model retrieval,mesh deformation and other operations.Nowadays,triangular mesh segmentation technology is widely used in industrial manufacturing,model design and other fields.In this dissertation,we further study the triangular mesh segmentation and some geometric problems.1.An optimal sampling algorithm for seed points in triangulation: Aiming at the problems of over segmentation,sparse segmentation and boundary error in traditional triangular mesh segmentation,we detect them respectively and correct them in three ways: inter class merging,intra class re clustering and boundary re dividing,so as to optimize the results of triangular mesh segmentation and extract high-quality seed points.Experiments show that the algorithm is used to extract seed points,generate contour lines and reconstruct models Better results.2.A rule-based block estimation method for triangular mesh model: Fitting the original mesh model by plane,cylinder,cone and sphere respectively.Find the best way and number of blocks,split the original model into several similar surfaces,and calculate the best number of blocks.Finally,the optimal number of blocks of the whole model is obtained.Finally,Proved its correctness by experiment.3.System implementation and application examples: The system is mainly used for the optimal sampling of seed points,model block estimation,and other 3D model geometry processing.And through an example of an application scenario to further prove the value of this article.In this thesis,we propose a method to solve the two kinds of geometric processing problems,which are seed point sampling and model block estimation in triangulation.The results of this thesis are integrated into a system software,and the practical value of this study is verified by an application example.