摘要
在模糊C-均值聚类算法的基础上,提出了基于改进粒子群和模糊C-均值聚类的混合图像分割算法.该算法利用改进粒子群算法优化模糊C-均值的目标函数,同时引入聚类有效性指标,通过迭代更新搜索到合理的分割类别数和聚类中心实现自动确定图像分割最佳类别数,并根据最佳类别数确定最优聚类中心的选取,最终实现图像的自适应分割.实验结果表明,该方法可自适应地确定图像分割最佳类别数,并能快速准确地实现图像分割.
A novel algorithm is presented for image segmentation based on improved particle swarm optimization (PSO) and fuzzy C-means clustering algorithm. The new algorithm optimizes fuzzy C-means (FCM) objective function through using the improved PSO. The proposed algorithm introduces validity index of cluster and searches for an appropriate num- ber of segmentation species and clustering center to determine the optimal clustering number of image segmentation automatically. The optimal clustering center can be decided to realize image segmentation according to the optimal clustering number. The experimental results indicate that the proposed algorithm automatically determines the optimal clustering number of image segmentation, meanwhile this method facilitates fast and correct image segmentation.
出处
《河北工业大学学报》
CAS
北大核心
2011年第6期6-10,共5页
Journal of Hebei University of Technology
基金
河北省科技支撑计划(10213565)
河北省高等学校科学技术研究指导项目(Z2010232)
关键词
粒子群优化算法
模糊C_均值聚类
聚类中心
有效性指标
图像分割
particle swarm optimization
fuzzy C-means clustering
clustering center
validity index
image segmentation