基于蚁群模糊聚类算法的图像边缘检测

提出了一种基于蚁群动态模糊聚类算法的图像边缘检测,该算法首先利用蚁群算法的较强处理局部极值的能力,克服了FCM算法对初始化的敏感,动态地确定了聚类数目和中心;然后利用蚁群聚类得到的结果,再进行FCM聚类弥补蚁群算法的不足.两者有机结合起来可以寻求到具有全局分布特性的最优聚类,实现了基于改进的目标函数聚类分析.最后将该算法应用到图像边缘检测,对比实验表明,该算法具有很强的模糊边缘和微细边缘检测能力.

Fuzzy clustering analysis based on ant colony algorithm for image edge detection

This paper proposes a method of dynamic fuzzy clustering analysis based on ant colony algorithm for image edge detection.The algorithm makes use of the great ability of ant colony algorithm for disposing local extremum firstly,which overcomes sensitivity to initialization of fuzzy clustering method(FCM) and fixes on the numbers of clustering as well as the centers of clustering dynamically.And then the results from previous for fuzzy clustering method can make up the deficiency of ant colony algorithm.In this way,we combine ant colony algorithm with fuzzy C-means clustering organically and find the whole distributing optimization clustering and achieve clustering analysis based on improved function.At last,the application of the algorithm proposed to image edge detection and comparative experiments show that the algorithm have great ability of detection the fuzzy edge and exiguous edge.