计算机工程与应用2012，Vol.48Issue(32)：164-169,6.DOI:10.3778/j.issn.1002-8331.1111-0197
空间模糊C均值聚类的神经切片图像分割方法
Spatial information based fuzzy C-means for nerve sliced image segmentation
摘要
Abstract
Peripheral nerve sliced microscopic images have the characteristics of complex background, discontinuous regions and non-uniform illumination.It is difficult to apply classical image segmentation algorithms to obtain a valid segmentation result.By combining the probability of the initial membership functions and space distance to design the space function of SFCM clustering algorithm, it proposes SFCM color image segmentation method in this paper.The image color space is converted from RGB to HIS color space.Cluster validity function is used to define the number of clusters of each component and initialize the algorithm with the algorithm of FCM based on image histogram.SFCM is applied separately on each component of HSI model, and each component is combined and displayed in RGB model.The experimental results show that compared with the standard FCM clustering segmentation algorithm, the new method is more effective for segmenting discontinuous nerve sliced microscopic images.关键词
模糊C均值聚类/空间模糊C均值聚类/彩色图像分割/神经切片/显微图像Key words
fuzzy C-means/ spatial fuzzy C-means/ color image segmentation/ nerve slice/ microscopic image分类
信息技术与安全科学引用本文复制引用
邹继杰,唐平,张毅,罗鹏,江小平,汪婷..空间模糊C均值聚类的神经切片图像分割方法[J].计算机工程与应用,2012,48(32):164-169,6.基金项目
广东省自然科学基金项目(No.9151008901000006) （No.9151008901000006）
广州市科技计划项目(No.12C22111580). （No.12C22111580）
