一种基于区域模糊特征的图像检索方法  

A Region-based Image Retrieval with Fuzzy Features Method

作　　者：唐敏[1] 姜灵敏[2] 阳爱民[1] 

机构地区：[1]湖南工业大学计算机科学与技术系,湖南株洲412008 [2]广东外语外贸大学信息科学技术学院,广州510420

出　　处：《郑州大学学报（理学版）》2007年第2期122-127,共6页Journal of Zhengzhou University:Natural Science Edition

基　　金：湖南省自然科学基金资助项目;编号05JJ40101

摘　　要：提出一种新的基于区域模糊特征的图像检索方法(Region-Based Image Retrieval with Fuzzy Features,RBIRFF).先将图像分割成区域后,利用模糊集理论计算每个区域的模糊特征,再利用区域特征间相似度计算方法,进行图像相似度测量.最后,按照相似度的降序排列图像,得到检索结果.介绍了RBIRFF原型系统的设计和实现,并用实验测评了其检索性能.提出的方法对大型图像数据库的快速和准确检索,有广泛的应用前景.A novel method, called region-based image retrieval with fuzzy features（RBIRFF）, is introduced, which partitions image into regions firstly, then computes each region fuzzy features with fuzzy set theory, and computes similarities of images. Finally, the retrieval results are obtained. The design and realization of RBIRFF system are also proposed, and its retrieval effectiveness is validated by experiments. This method for image databases efficiency and effectiveness has a wide application.

关 键 词：CRIR 模糊集理论 模糊特征 相似度计算 

分 类 号：TP39[自动化与计算机技术—计算机应用技术]