图片模糊相似度量及其在模式识别中的应用

Picture fuzzy similarity measure and its application in pattern recognition

目的:研究图片模糊集上一种新的相似度量。方法:首先通过权重系数将拒绝隶属度携带的信息进行再分配。其次,利用聚合函数与取大取小算子构造了图片模糊集上的相似度量。结果:提出的相似度量不仅满足相似度量的公理化定义,而且克服了现有的一些相似度量的某些不足。结论:数值实验说明了我们提出的图片模糊集上的相似度量是合理的,我们进而将提出相似度量应用到模式识别中,其识别结果表明我们的相似度量是有效的优越的。

Aims:This paper aims to study new similarity measures for picture fuzzy sets.Methods:The information carried by the refusal membership degree was redistributed through the weight coefficient.The similarity measure was constructed on the picture fuzzy set by using the aggregation function.Results:The proposed similarity measure not only satisfied the axiomatic definition of similarity measure,but also overcame some shortcomings of existing similarity measures.Conclusions:Numerical experiments show that our similarity measure is reasonable.Further,we applied the proposed similarity measure to pattern recognition;and the recognition results showed that our similarity measure was effective and superior.