基于蚁群及空间邻域信息的FCM图像分割方法  

Image Segmentation Based on the Ant Colony and Improved FCM Clustering Algorithm with Spatial Information

作　　者：毛晓波[1] 张勇杰[1] 陈铁军[1] 

机构地区：[1]郑州大学电气工程学院,河南郑州450001

出　　处：《郑州大学学报（工学版）》2014年第1期1-4,共4页Journal of Zhengzhou University（Engineering Science）

基　　金：高等学校博士学科点专项科研基金资助项目(20114101110005);河南省重大科技攻关项目(102101210100);河南省教育厅科学技术研究重点项目资助计划项目(12A410002)

摘　　要：针对模糊C均值(FCM)聚类算法聚类个数难以确定、搜索过程易陷入局部最优的缺陷,把蚁群算法与改进的FCM聚类算法相结合,提出了一种基于蚁群算法的带有空间邻域信息的模糊C均值聚类图像分割算法.首先利用分水岭算法对图像进行初始分割,然后利用蚁群算法寻优,求得聚类中心和聚类个数,将其作为模糊C均值聚类的初始聚类中心和聚类个数进行模糊聚类.实验结果表明:由于聚类样本数量显著减少,很大程度上提高了聚类速度和抗噪能力,增强了算法的鲁棒性.With the fuzzy Cmeans clustering （FCM） algorithm it is difficult to determine the number of clus ters on image segmentation, which is easy to get into a local optimum. In order to solve the problems, this pa per proposed a new segmentation method based on the ant colony and improved FCM Clustering Algorithm with spatial information. Dividing image with the help of watershed algorithm, we got the initial segmentation re sults. It made full use of the ability of global optimization of the ant colony algorithm to obtain the accurate o riginal cluster centers and cluster number. Then the results were obtained as the initial cluster centers and the number of clusters of fuzzy Cmeans clustering algorithm. The experimental results show that： due to the de crease of the size of clustering samples, the clustering speed, noise immunity and the robustness of the algo rithm are improved significantly.

关 键 词：蚁群算法 分水岭 空间约束 图像分割 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]