基于Gibbs随机场与模糊C均值聚类的图像分割新算法

模拟C均值聚类(FCM)是一种非常经典的非监督聚类技术,已被广泛用于图像的自动分割.由于传统的FCM算法进行图像分割仅利用了灰度信息,而没有考虑象素的空间位置信息,因而分割模型是不完整的,造成传统FCM算法只适用于分割噪声含量很低的图像.为了克服传统FCM算法的局限性,本文利用Gibbs随机场所描述的邻域关系属性,引入先验空间约束信息,提出拒纳度的概念,建立包含灰度信息与空间信息的新聚类目标函数,继而提出基于Gibbs随机场与模糊C平均聚类的GFCM图像分割新算法.实验证明,利用本文所提GFCM算法可以有效地分割含噪声图像.

A New Algorithm for Image Segmentation Based on Gibbs Random Field and Fuzzy C-Means Clustering

Fuzzy c-means(FCM) clustering is one of well-known unsuperviaed clustering techniques, which has been widely used in automated image segmentation.However, when the classical FCM algorithm is used for image segmentation, no spatial information is taken into account.This causes the FCM algorithm to work only on well-defined images with low level of noise;unfortunately, this is not often the case in reality. In order to overcome this limitation of FCM, the prior spatial constraint is incorporated based on Gibbs random field theory.The definition of re/usable level is presented and then new clustering object function is presented.This new algorithm connects Gibbs random field with FCM algorithm and is shown to be most effective in our experiments.