摘要
提出一种基于Sugeno模糊积分的模糊相似度量并用于图像检索中。文章用模糊测度来描述人的主观反馈。试验表明该文的方法可以大大提高图像检索系统的效率和稳定性,在反馈后的表现要优于加权平均方法(WAO)和采用Choquet积分(CI)的方法。
In this paper,we propose a fuzzy similarity measure based on Sugeno fuzzy integral and apply it to Content-Based Image Retrieval.Fuzzy measure is used to reflect human subjective feedback.Experimental results show that the proposed method can improve the efficiency and the stability of image retrieval system.The behavior of the method in feedback is better than WAO(weighted average operator)and Choquet Integral-based similarity measure.
出处
《计算机工程与应用》
CSCD
北大核心
2005年第2期36-39,74,共5页
Computer Engineering and Applications
基金
国家自然科学基金(编号:60302012)
宁波市重点博士基金(编号:2003A61006)
关键词
图像检索
加权平均
模糊测度
模糊积分
image retrieval,weighted average operator,fuzzy measure,fuzzy integral
