摘要
如何构建有效的组织和索引、提高图像检索速度是基于内容的图像检索所需解决的关键问题之一。论文采用了一种基于改进的模糊C均值算法的聚类索引。实验表明:该方法应用于图像检索,在准确性和实时性方面均能达到较好的效果,并优于已有的模糊C均值聚类算法。另外,系统实现了基于多特征结合的方法进行检索,并利用基于相关反馈的权重调整方法进一步提高检索性能,使检索结果更加符合用户的视觉效果。
One of the most important issues in content-based image retrieval(CBIR)is how to construct effective orga-nization and index to enhance image retrieval speed.Clustering is a kind of effective indexing method.This paper proposes a modified fuzzy C-means(MFCM)clustering algorithm to construct index of the entire images database before retrieval.Experiments show that MFCM applied to image retrieval is effective in exactness and real-time property.It is superior to traditional fuzzy C-means clustering algorithm.In addition,it uses multi-features weight adjusting method to improve the performance of the system,the result of retrieval will satisfy people's visual receptance.
出处
《计算机工程与应用》
CSCD
北大核心
2004年第31期46-48,共3页
Computer Engineering and Applications
基金
国家自然科学基金项目(编号:59778050)资助