摘要
针对模糊聚类图像分割算法的固有缺点,提出了一种基于D-S证据理论的模糊聚类图像融合分割算法。对图像的点灰度特征和块灰度特征分别进行模糊C均值聚类,并将各自的模糊隶属度转化为单一或复合假设及其基本概率赋值,再利用D-S证据理论进行融合分割。实验结果表明该算法的分割效果优于传统的模糊聚类分割算法。
In view of the defects of fuzzy C-mean clustering(FCM) image segmentation,an FCM image fusion segmentation algorithm based on the Dempster-Shafer(D-S) theory is presented. Based on the clustering results of the gray level feature of each pixel and their spatial pixels,each of the membership degree is translated into the basic belief assignment of simple or composite hypotheses,then these basic belief assignments are combined to make fusion segmentation according to the D-S theory. The experiment results demonstrate the excellent performance of this algorithm.
出处
《合肥工业大学学报（自然科学版）》
CAS
CSCD
2004年第7期721-724,共4页
Journal of Hefei University of Technology：Natural Science
基金
国家自然科学基金资助项目(60175011
60375011)
安徽省自然科学基金资助项目(01042301)
安徽省重点科研基金资助项目(03021012)
