摘要
提出一种基于法向距离的法矢方向调整算法。主要思想是根据法向距离阈值,把散乱点划分为平坦点和非平坦点两种类型;根据其邻近点是否有不平坦点来采用不同的法矢传播方式而无需建立散乱点法矢的Riemannian图;并提出了两次最近距离和一次k近邻遍历法加快了法矢的传播速度。应用实验球、切牙、磨牙、怪兽和全口牙模5个不同点云数据进行了算法验证,对怪兽和全口牙模法矢传播过程中遇到的死锁问题,使用三次最近距离法给予了解决。
We divide the scattered points into planar points and non-planar points according to the threshold value of a given normal distance. Depending on whether the k neighbor has non-planar points, we decide the ways of normal vector spreading, thus making it unnecessary to establish the Riemannian graph of the scattered points' normal vector. Moreover, to accelerate the normal vector spreading, we propose the two-order nearest distance and one-order k-neighbor method. We verify the algorithm with the five types of scattered point-cloud data from experimental sphere, incisor, molar, monster and full teeth model. We adopt the three-order nearest distance method to solve the dead-lock problem during the spreading of normal vectors of monster and full teeth model.
出处
《机械科学与技术》
CSCD
北大核心
2008年第2期192-197,共6页
Mechanical Science and Technology for Aerospace Engineering
基金
国家高技术研究发展计划(863计划)项目(2005AA420240)
江苏省科技攻关项目(BE2005014)
南京市医学科技发展计划项目(ZKX0420)
南京市科技发展计划项目(200504022)
江苏省自然科学基金项目(BK2006060)资助
关键词
曲面重建
散乱点云
切平面法矢
后近邻
法矢传播
surface reconstruction
scattered point
normal vector spreading
k-neighbor
spreading of normal vector