基于区域合并的FCM图像分割改进算法

Improved FCM Image Segmentation Algorithm Based on Region Merging

针对现有模糊C均值聚类(FCM)算法易出现过分割现象,分割效果不够理想等问题,本文提出了一种基于区域合并的FCM改进算法.该算法首先使用快速广义模糊C均值聚类算法(FGFCM)获得初始分割;然后综合考虑各区域间的邻接关系、颜色差异和边缘信息,计算各邻接区域间的距离;最后依据区域间距离和区域面积对初始分割区域进行合并,得到最终分割结果.实验证明,所提出的算法有更好的分割性能,有效解决了现有FCM分割算法中的过分割问题.

Aiming at the problems of over-segmentation and unsatisfactory segmentation in the existing fuzzy C-means clustering （FCM） algorithms，an improved FCM algorithm based on region merging is proposed in this paper.Firstly，the initial segmentation is obtained by the fast-generalized fuzzy C-means clustering algorithm （FGFCM）.Then，the distance between adjacent regions is calculated by taking into account adjacency relation，color difference and edge information.Finally，the final segmentation result is obtained by merging the initial segmentation region based on the distance and area.Experiments show that the proposed algorithm not only has better segmentation performance，but also effectively solves the problem of oversegmentation in existing FCM segmentation algorithm.