Texture image segmentation using Brushlet-domain hidden Markov models

Fang Liu; Kai Yang; Hongxia Hao; Biao Hou; Hua Zhong

doi:10.1117/12.833002

30 October 2009 Texture image segmentation using Brushlet-domain hidden Markov models

Fang Liu, Kai Yang, Hongxia Hao, Biao Hou, Hua Zhong

Author Affiliations +

Proceedings Volume 7498, MIPPR 2009: Remote Sensing and GIS Data Processing and Other Applications; 749854 (2009) https://doi.org/10.1117/12.833002
Event: Sixth International Symposium on Multispectral Image Processing and Pattern Recognition, 2009, Yichang, China

Abstract

By researching the Brushlet domain coefficients of texture images, we found that the distribution of the magnitudes of Brushlet domain coefficients roughly meet rayleigh distribution. And there are correlations between Brushlet coefficients in adjacent scales. Therefore, Rayleigh Mixture Model (RMM) is used to characterize the statistics of the magnitudes of Brushlet coefficients. To capture the inter-scale persistence of Brushlet coefficients, a "four to four" models with markov property is adopted in this paper. On the basis, by combining with the multi-scale Bayesian segmentation method, we propose a multiscale Bayesian texture segmentation algorithm that is based on a Brushlet domain hidden Markov tree (BruHMT) model. The experiment results indicate that our method is feasible and effective. Especially for coarse texture, our method is superior than texture segmentation method using Wavelet domain hidden Markov tree (WHMT) model.

Citation Download Citation

Fang Liu, Kai Yang, Hongxia Hao, Biao Hou, and Hua Zhong "Texture image segmentation using Brushlet-domain hidden Markov models", Proc. SPIE 7498, MIPPR 2009: Remote Sensing and GIS Data Processing and Other Applications, 749854 (30 October 2009); https://doi.org/10.1117/12.833002

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

;

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE