摘要
针对三角网格模型中的二次曲面提取问题,利用随机抽样一致性算法(Random Sampling Consensus,RANSAC)提出了一种球面、圆柱面和圆锥面等二次特征曲面的提取算法。该算法首先计算每个顶点的法向量,然后用带顶点法向量的三角面片统一表示每类二次曲面的最小子集,最后利用RANSAC算法进行二次曲面提取。实验结果分析表明该算法具有良好的稳定性和运算效率等特点。
An automatic algorithm was presented to detect quadric from triangular mesh. The method is based on random sampling consensus algorithm and iteratively detects planes, spheres, cylinders and cones. In order to reduce the number of required points to define a unique instance of each quadric, the algorithm computes an approximate surface normal for each point firstly, then the minimal subset is defined as a triangle with local orientation to improve the robust of the algorithm. Experimental results demonstrate that the algorithm is capable to extract a variety of different types of quadric, while retaining such favorable properties as high robustness, effectiveness and generality.
出处
《机床与液压》
北大核心
2009年第1期14-16,8,共4页
Machine Tool & Hydraulics
基金
总装备部预先研究项目资助(41318.1.1.7)
航空科学基金资助项目(2008ZB55001)
关键词
随机抽样一致性算法
三角网格
二次曲面提取
Random sampling consensus algorithm
Triangular mesh
Quadric surface extraction
