Computer Engineering and Applications ›› 2020, Vol. 56 ›› Issue (4): 197-201.DOI: 10.3778/j.issn.1002-8331.1811-0222

Previous Articles     Next Articles

Improved FCM Image Segmentation Algorithm Based on Markov Random Field

WANG Yan, QI Xianghui, DUAN Yaxi   

  1. College of Computer and Communication, Lanzhou University of Technology, Lanzhou 730050, China
  • Online:2020-02-15 Published:2020-03-06



  1. 兰州理工大学 计算机与通信学院,兰州 730050


Aiming at the problem that the utilization rate of neighborhood information and spatial information of fuzzy clustering algorithm is low and susceptible to noise, an improved Fuzzy C-Means(FCM) algorithm based on Markov random field model, namely FKMFCM algorithm, is proposed. On the basis of FCMKM algorithm, the prior probability of Markov random field is added, and the objective function of FCM algorithm is improved by using the prior probability. In order to verify the performance of FKMFCM algorithm, Bezdek partition coefficient, Xie_Beni coefficient, running time and iteration times are selected as the evaluation criteria of comparative experiments. The experimental results show that the FKMFCM algorithm can effectively improve the anti-noise performance of the fuzzy clustering algorithm.

Key words: Markov random field, image segmentation, fuzzy clustering, neighborhood information, noise resistance


针对模糊聚类算法邻域信息与空间信息利用率低,易受噪声影响的问题,提出一种结合马尔科夫随机场模型的改进模糊C均值算法(Fuzzy C-Means,FCM),即FKMFCM算法。在FCMKM算法基础上添加马尔科夫随机场先验概率,利用先验概率改进FCM算法的目标函数,提高FCM算法抗噪性。为验证FKMFCM算法的性能,选取Bezdek划分系数、Xie_Beni系数、运行时间、迭代次数4个评测指标作为对比实验的评价标准。实验结果表明,FKMFCM算法能有效地提高模糊聚类算法的抗噪性。

关键词: 马尔科夫随机场, 图像分割, 模糊聚类, 邻域信息, 抗噪性