运用边界状态约束的表面体素加密细分算法  

Surfacevoxel refinement algorithm using boundary state constraints

作　　者：刘晨燕 敬石开[1] 张伟[1,2] 赵芳垒[1] LIU Chenyan;JING Shikai;ZHANG Wei;ZHAO Fanglei(School of Mechanical Engineering,Beijing Institute of Technology,Beijing 100081,China;Institute of Computing Technology,Chinese Academy of Science,Beijing 100190,China)

机构地区：[1]北京理工大学机械工程学院,北京100081 [2]中国科学院计算技术研究所,北京100190

出　　处：《计算机集成制造系统》2020年第4期970-979,共10页Computer Integrated Manufacturing Systems

基　　金：国家863计划资助项目(2015AA042505)。

摘　　要：为了平衡体素模型的体素数量与精细度,提出一种实体模型体素加密细化算法。该算法对低分辨率均一体素模型的表面体素进行切分,标记与三角面不相交的体素,并运用边界状态传递判定标记体素的内外位置。通过编码不同八叉树的节点,并进行跨八叉树快速邻居节点搜寻,实现多级加密细分。实验结果表明,该算法可以以低分辨率模型为基础,精确地获得体素单元少的高分辨率模型。To balance the voxel number and surface fineness of voxel model,a refinement method for solid models was proposed.The surface voxels of the low-resolution uniform model were divided firstly,and sub-voxels not intersecting any triangle were marked.Boundary state constraints method was used to determine whether a marked voxel was inside or outside the model.Neighbor searching scheme across different octree nodes was achieved by encoding rules,which could realize multi-level refinement.The result showed that starting from the low-resolution voxel model,the proposed method could produce accurate high-resolution model with fewer voxels.

关 键 词：表面细分 多分辨率 体素化 边界状态约束算法 

分 类 号：TP391.7[自动化与计算机技术—计算机应用技术]