一种保持尖锐特征的3^(1/2)细分算法

A 3^(1/2)-Subdivision Algorithm for Holding Sharp Features

提出一种能够重建模型尖锐特征的3细分算法.首先根据预设的阈值θ自动标记网格中的尖锐特征边,并计算出各个顶点的尖特征度;然后根据顶点尖特征度的不同修改原3细分方法的几何规则,设计出相应的权值掩模(mask);最后通过在奇数次细分时不翻转特征边,在偶数次细分时插入边点的方法来实现尖锐特征的重建.实验结果表明,与原3细分方法相比,该算法能够更好地保持模型的尖锐特征.

This paper presents a √3-subdivision algorithm to rebuild sharp features. First, sharp edges are automatically tagged according to the given threshold θ and the sharp degree of each vertex is computed. Then, based on different sharp degrees, geometric rules of √3-subdivision scheme are modified and subdivision masks are designed. Finally sharp features are reconstructed by keeping sharp edges non-flipping in odd refinement steps and inserting edge points in even refinement steps. Experimental results show that this algorithm keeps sharp features better than original √3-subdivision method does.