Application of Improved Fuzzy Clustering Method in the Image Segmentation

摘要：

This paper presents an operator of fuzzy clustering method of image segmentation based on Local Binary Pattern (LBP). Semi-supervised learning and fuzzy clustering method are introduced in order to overcome the problem of initial clustering sensitive. Also, local binary pattern operator is introduced to construct the space feature vectors of pixels, which makes full use of the space characteristics information of pixel. Space feature vectors of pixels are used to choose image segmentation clustering data set. Result of experiments shows that the method improves the accuracy of segmentation and enhance the detailed features of segmentation.

关键词： Semi-supervision fuzzy clustering Local value of the binary mode Initial clustering center

作者: Lijun Peng Lingmin He Xiaobing Yang Kangjian Wang

作者单位: College of Information Engineering, China Jiliang University, Zhejiang Province, China

会议类型: 国际会议

会议名称: 2012 Fifth International Symposium on Computational Intelligence and Design 第五届计算智能与设计国际会议 ISCID 2012

会议地点: 杭州

会议语种:英文

页码: 639-642

在线出版日期: 2012-10-28（万方平台首次上网日期，不代表论文的发表时间）