Abstract: To improve the efficiency of triangular mesh model surface voxelization algorithms,a fast surface voxelization algorithm was proposed based on triangle subdivision.In this algorithm,a new triangle subdivision method was adopted to obtain a large number of vertexes information,the threshold value of triangle area was used to control the number of generated vertexes and the the point vector translation method was used reduce the computational resource of vertex coordinates.Through building the corresponding relationship between triangle vertexes and voxel units,the model surface voxelization was realized.A variety of model instances with different resolutions were used to compare with two existing algorithms.The results showed that the number of needed sampling points of surface voxelization was less than the existing algorithms and the surface voxelization efficiency proposed algorithm was superior to existing algorithms.The algorithm was also applicable to the mesh model whose genus was not zero and mesh model with inner closed surface.

Key words: surface voxelization, triangular mesh model, triangle subdivision, vector translation

摘要： 针对现有三角网格模型表面体素化算法效率低的问题,提出一种基于三角形细分的三角网格模型表面体素化算法。该算法采用三角形细分方式获取大量顶点信息,采用面积阈值控制顶点的生成数量,并采用点的向量平移方式减少求解顶点坐标的计算量。通过建立顶点信息与三角形相交体素单元的一一对应关系实现三角网格模型的表面体素化。通过与现有2种算法在不同分辨率下的多种模型实验对比,表明该算法所需的采样点数量少于现有算法,在模型表面体素化效率方面优于现有算法。此外,所提算法也适用于亏格不为0和存在封闭内腔的三角网格模型。

关键词: 表面体素化, 三角网格模型, 三角形细分, 向量平移