基于变分高斯混合模型的图像分割算法(英文)

被引:1
作者
张媛媛
钟意伟
机构
[1] 宁波大学信息科学与工程学院
关键词
图像分割; 变分推断; 高斯混合模型; 期望最大化算法;
D O I
暂无
中图分类号
TP391.41 [];
学科分类号
080203 ;
摘要
提出了一种基于变分推断的高斯混合模型的图像分割算法.该算法首先用贝叶斯混合高斯模型对图像的特征进行建模,并针对模型的参数学习问题,利用变分推断算法估计模型的参数及其后验概率;这种方法比采样法的计算量更少,而且能够根据图像数据自动优化混合个数,实现了模型的自动选择.最后,该算法在Berkeley的自然图像集上进行的实验结果与经典的图像分割算法进行了比较,结果表明此方法得到的图像分割结果精度较高,具有较好的性能.
引用
收藏
页码:23 / 28
页数:6
相关论文
共 15 条
[1]  
"Variational inference for bayesian mixtures of factor analysers,". Z. Ghahramani,M. Beal. Advances in NeuralInformation Processing Systems . 2000
[2]  
Pattern Recognition and Machine Learning. Christopher M Bishop. . 2007
[3]  
Combining region and edge cues for image segmentation in a probabilistic Gaussian mixture framework. Omer R,Hayit G,Jacob G. Computer Vision and Pattern Recognition . 2007
[4]  
Performance of Bayesian model selection criteria for Gaussian mixture models. Steele R J,Raftery A E. . 2009
[5]  
Variational Bayesian inference with stochastic search. John P,David B,Jordan M I. Proceedings of the29th International Conference on Machine Learning . 2012
[6]  
Stochastic variational inference. Hoffman M,Blei D M,Wang Chong,et al. Journal of Machine Learning Research . 2012
[7]  
Graphical models, exponential families, and variational inference. Wainwright, Martin J.,Jordan, Michael I. Foundations and Trends in Machine Learning . 2008
[8]  
A spatially constrained mixture model for image segmentation. Blekas, K.,Likas, A.,Galatsanos, N.P.,Lagaris, I.E. IEEE Transactions on Neural Networks . 2005
[9]  
ON IMAGE SEGMENTATION USING INFORMATION THEORETIC CRITERIA. Alexander Aue,Thomas C. M. Lee. The Annals of Statistics . 2011
[10]  
Unsupervised learning of finite mixture models. Figueiredo M A T,Jain A K. IEEE Transactions on Pattern Analysis and Machine Intelligence . 2002