基于细节的自适应网格简化  

Detail-Preserving Adaptive Mesh Simplification

作　　者：高山[1] 卢汉清[1] 周万宁[2] 

机构地区：[1]中国科学院自动化研究所模式识别国家重点实验室,北京100080 [2]海军论证中心自动化研究所,北京100036

出　　处：《计算机辅助设计与图形学学报》2003年第9期1122-1127,共6页Journal of Computer-Aided Design & Computer Graphics

基　　金：国家自然科学基金 ( 6980 5 0 0 5 )资助

摘　　要：从模型中保留的几何细节出发 ,提出一种自适应的三角网格简化算法 该算法首先比较顶点与其相邻点之间的欧氏距离是否超过预先设定的简化尺度 ,删除小于该尺度的相邻点 ,再对删除造成的空洞作局部三角剖分 此算法的特点是用细节度参数控制简化模型的整体精度 ,并且简化的尺度可以根据模型表面的细节情况作自适应调整 ,自动在细节丰富的区域变小而在细节稀疏的区域变大 通过两组应用实例可以看到 。From the point of view of keeping geometric details, we present an adaptive simplification method for triangular mesh. We first compute the distance between a vertex and its neighbor, then delete the neighbor within the local simplifying scale and triangulate the hole caused by deletion. Here, the local simplifying scale is the ratio of the local feature size(LFS) to a given global precision control parameter. Amenta N. and Bern M. pointed out that the smaller LFS is, the more details appear on the area where the vertex is located. Therefore, the local simplification scale is inversely proportional to the global precision control parameter; and it also automatically adjusts according to the surface of model with a given global precision control parameter. The results show that this method can not only reduce the amount of data effectively but also preserve the vision characters very well.

关 键 词：自适应网格简化 计算机图形学 网格模型 多边形网格模型 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]