3D Mesh Smoothing Method Based on Vertex Classification and Distance Equalization

Abstract

The triangular mesh model obtained from the measured data in reverse engineering often contains a large amount of noise, disturbance and irregular triangles, which need to be smoothed to meet the requirements of subsequent processing. This paper analyzes the advantages and disadvantages of Laplacian smoothing method and Taubin method in terms of processing effect and operation efficiency. A smoothing method based on vertex classification and distance equalization is proposed in this paper to avoid the drawbacks of Laplace smoothing method and Taubin method. Vertex classification divides the vertices in the mesh model into feature vertices and normal vertices, so that different smoothing methods can be used for different vertices. The smoothing method based on distance equalization starts from the distance between the vertices of the mesh and the average plane of its neighbors, and takes the difference of the average plane distance from the adjacent vertices to its neighborhood as the weight to control the moving direction of the vertices, effectively maintaining the surface features of the model.