摘要
提出一种基于空间邻域信息的FCM图像分割算法,该方法将目标函数中的距离定义为特征距离与空间距离之和,不仅反映特征距离,而且反映空间距离.将空间信息引入到传统FCM算法的目标函数中,建立了包含邻域信息的新的聚类目标函数,实现图像的分割.实验结果表明,新算法能够获得较好的分割效果和质量,同时具有较强的抑制噪声的能力.
A novel fuzzy C-means algorithm based on the spatial information is proposed, where the distance is defined as a combination of the Euclidean distance in the feature space and the spatial distance, which not only reflects the feature distance, but also reflects the spatial distance. A novel objective function has been established which contains neighbor information. Experimental results indicate that the proposed algorithm can get better segmentation result and is more robust to noise.
出处
《华中科技大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2009年第6期56-59,共4页
Journal of Huazhong University of Science and Technology(Natural Science Edition)
基金
国家自然科学基金资助项目(60574025)
湖北省自然科学基金资助项目(2004ABA055)
关键词
图像分割
分割算法
模糊C均值
邻域信息
鲁棒性
image segmentation
segmentation algorithm
fuzzy C-means
neighbor information
robust
