DocumentCode :
932294
Title :
Image Segmentation Using Hidden Markov Gauss Mixture Models
Author :
Pyun, K.P. ; Johan Lim ; Chee Sun Won ; Gray, R.M.
Author_Institution :
Stanford Univ., Stanford
Volume :
16
Issue :
7
fYear :
2007
fDate :
7/1/2007 12:00:00 AM
Firstpage :
1902
Lastpage :
1911
Abstract :
Image segmentation is an important tool in image processing and can serve as an efficient front end to sophisticated algorithms and thereby simplify subsequent processing. We develop a multiclass image segmentation method using hidden Markov Gauss mixture models (HMGMMs) and provide examples of segmentation of aerial images and textures. HMGMMs incorporate supervised learning, fitting the observation probability distribution given each class by a Gauss mixture estimated using vector quantization with a minimum discrimination information (MDI) distortion. We formulate the image segmentation problem using a maximum a posteriori criteria and find the hidden states that maximize the posterior density given the observation. We estimate both the hidden Markov parameter and hidden states using a stochastic expectation-maximization algorithm. Our results demonstrate that HMGMM provides better classification in terms of Bayes risk and spatial homogeneity of the classified objects than do several popular methods, including classification and regression trees, learning vector quantization, causal hidden Markov models (HMMs), and multiresolution HMMs. The computational load of HMGMM is similar to that of the causal HMM.
Keywords :
hidden Markov models; image classification; image segmentation; bond-percolation model; hidden Markov gauss mixture models; image classification; image processing; image segmentation; maximum a posteriori criteria; minimum discrimination information distortion; observation probability distribution; parameter estimation; stochastic expectation-maximization algorithm; supervised learning; vector quantization; Classification tree analysis; Gaussian distribution; Gaussian processes; Hidden Markov models; Image processing; Image segmentation; Probability distribution; State estimation; Supervised learning; Vector quantization; 2-D hidden Markov models (HMMs); Bond-percolation (BP) model; Gauss mixture models (GMMs); Gauss mixture vector quantizer (GMVQ); image classification; image segmentation; parameter estimation; Algorithms; Artificial Intelligence; Computer Simulation; Image Enhancement; Image Interpretation, Computer-Assisted; Markov Chains; Models, Statistical; Normal Distribution; Pattern Recognition, Automated;
fLanguage :
English
Journal_Title :
Image Processing, IEEE Transactions on
Publisher :
ieee
ISSN :
1057-7149
Type :
jour
DOI :
10.1109/TIP.2007.899612
Filename :
4237207
Link To Document :
بازگشت