基于模糊C均值聚类的织物平整度等级评定  

FABRIC WRINKLE GRADE ASSESSMENT BASED ON FUZZY C MEDIAN CLUSTER

作　　者：杨晓波[1] 

机构地区：[1]浙江财经学院信息学院,浙江杭州310012

出　　处：《计算机应用与软件》2006年第9期43-44,50,共3页Computer Applications and Software

基　　金：浙江省教育厅科研项目(编号:20041420)

摘　　要：采用模糊C均值聚类客观评定织物平整度等级。首先介绍了模糊C均值聚类的基本原理,模糊聚类可以将输入特征值进行聚类并分组;然后利用模糊C均值聚类对输入特征值进行聚类分析,不同平整度等级的织物模板被分属于不同的模糊聚类中心;最后选取26种不同类型的织物样本进行测试,试验结果表明,客观评价与主观评价的相关系数达到97.38%,评定准确率超过90%。Fuzzy C Median （ FCM ） cluster is proposed to evaluate the fabric wrinkle grade objectively. Firstly, the basic principle of FCM is introduced. FCM can divide the input feature value into several groups. Then, FCM is applied to analyze the input wrinkle feature value. Different wrinkle grade of fabric replica is attributed different fuzzy cluster center. Finally, twenty-six kinds of fabric sample are selected and tested. The result shows that the correlation coefficient between objective assessment and subjective assessment is up to 97.38%, and the accuracy ratio of assessment is over 90%.

关 键 词：模糊C均值聚类 平整度等级 模式识别 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]