摘要： 三维模型检索中的形状分布算法使用伪随机序列进行采样，导致检索精度较低、检索结果不稳定。该文采用拟随机序列中的Halton序列在[0, 1]区间内产生分布较均匀的随机数，利用数学变换将其转换为采样空间的随机数进行采样。三维模型检索对比实验结果表明，该方法可以提高检索精度，使检索结果更稳定。

关键词: 三维模型检索, 形状分布, 采样, 拟随机序列

Abstract: Shape distribution algorithm of 3D model retrieve uses pseudo random numbers to sample points on the model surface, which makes the retrieve precision low and the result unstable. This paper proposes a method which employs quasi random sequence to sample points on the model surface. More uniformly distributed random sequences in zone [0, 1] are generated based on Halton sequence, and transformed into the proper sequences in the sampling space. The 3D model retrieve experimental results show that the retrieve efficiency is improved by using this method.

Key words: 3D model retrieve, shape distributions, sampling, quasi random sequence

中图分类号: 

• TP391