高质量保形三角网格简化算法

High Quality and Shape Preserving Triangular Mesh Decimation

网格简化算法在多分辨率模型显示、三角网格模型重构、网络传输等领域应用广泛,但是目前的简化算法简化后的网格,经常会出现极大、极小、狭长的三角片,质量不高。在基于QEM(QuadricErrorMetric)简化方法的基础上,提出了在点对折叠代价中引入折叠点对关联形状因子、法向量的夹角因子、边长的非线性加权因子3种优化因子,并配合虚折叠回退操作,对简化过程中网格质量进行综合优化,并取得了较好的简化结果。

Triangular mesh decimation is widely used in multi-resolution model rendering, network transmission, computer animation and many other fields. Most of the existing algorithms for decimation of triangular meshes perform poorly at very low levels of detail. Abnormal triangles such as long-narrow, much larger or smaller than normal size are often introduced into the decimation results. A synthesis optimization method was proposed which considers the triangle shape, dihedral angle between two triangles which share a common edge and edge lengths in the process of calculating edge collapse costs based on QEM, and also propose a virtual edge collapse method attempting to find a local optimization. The proposed method can achieve much better regular decimation results under the condition of preserving visually important parts of the mesh details. The proposed method also has smaller memory consumption and execution time than most of the published notable algorithms.