摘要
通过引入采样定理,提出了基于视觉采样定理的新聚类算法,将生物物理学中的Weber定律、采样定理和视觉结构有效地结合起来,并在此算法的基础上提出了聚类的有效性准则.该算法基于视觉系统工作原理,具有更强的物理解释性能.实验表明,此算法简洁、有效.
Based on sampling theory, a new visual clustering algorithm was proposed. Combining the sampling theory, Weber law with clustering structure effectively, the main idea of this method is to consider the attractors corresponding to the differential functions of the sampling theory as the center of the cluster with the datum in the attraction domain as the members of the corresponding cluster. The simulations demonstrate the effectiveness of the new clustering method.
出处
《上海交通大学学报》
EI
CAS
CSCD
北大核心
2005年第4期544-548,共5页
Journal of Shanghai Jiaotong University
基金
中法先进计划项目(PRASI03-032)
关键词
聚类算法
视觉系统
采样定理
吸引子
Algorithms
Computer simulation
Data processing
Differential equations
Iterative methods
Sampling
Vision