3D Mesh Model Segmentation Method Based On Energy Optimization And Surface Features

As one of the cornerstones of the fourth generation of digital media,three-dimensional models have entered people's life and are widely used in many fields.Three-dimensional model segmentation is one of the key technologies of model processing,and has become a research hotspot of digital geometric processing.This paper elaborates on the advantages and disadvantages of the existing methods.Further research work has been done on the problem of unsatisfactory segmentation results in existing automated segmentation methods.This paper proposes mesh segmentation methods based on energy optimization and surface features.The main content is as follows:(1)Propose a mesh segmentation method based on energy optimization to solve the problems of inaccurate segmentation boundary and jagged segmentation of existing methods.On the basis of the energy function,the concept of difference is proposed,which use concavity,convexity and difference to measure segmentation.The segmentation boundary is found through the idea of corrosion algorithm and energy optimization,and Dijkstra's algorithm is used to optimize the segmentation boundary.The experiments show that the result of the proposed method is closer to the results of human manual segmentation,which improves the problems of inaccurate segmentation boundary and jagged segmentation of the existing methods.(2)Propose a CAD mesh segmentation method based on surface features to solve the problems of inaccurate segmentation boundary and meaningless results of existing methods.The proposed method takes normal,curvature,concavity,convexity as the main segmentation basis to decomposes the mesh surface into planes,concave and convex surfaces.Finally,according to the definition of region segmentation,it merges adjacent regions with specific rules to achieve meaningful segmentation of the CAD mesh.The proposed method achieves a better effect than prevailing methods.The segmentation boundary is more accurate,and the segmentation result has engineering significance.