MRF model-based algorithms for image segmentation
Richard C. Dubes, Anil Kumar Jain, Sateesha G. Nadabar et al. · Proceedings - International Conference on Pattern Recognition/Proceedings/International Conference on Pattern Recognition · 2002
The authors empirically compare three algorithms for segmenting simple, noisy images: simulated annealing (SA), iterated conditional modes (ICM), and maximizer of the posterior marginals (MPM). All use Markov random field (MRF) models to include prior contextual information. The comparison is based on artificial binary images which are degraded by Gaussian noise. Robustness is tested with correlated noise and with object and background textured. The ICM algorithm is evaluated when the degradation and model parameters must be estimated, in both supervised and unsupervised modes and on two real images. The results are assessed by visual inspection and through a numerical criterion. It is concluded that contextual information from MRF models improves segmentation when the number of categories and the degradation model are known and that parameters can be effectively estimated. None of the three algorithms is consistently best, but the ICM algorithm is the most robust. The energy of the a posteriori distribution is not always minimized at the best segmentation.>
Cited by 150 works. doi:10.1109/icpr.1990.118221
Authors
References
- Stochastic Relaxation, Gibbs Distributions, and the Bayesian Restoration of Images
- Spatial Interaction and the Statistical Analysis of Lattice Systems
- On the Statistical Analysis of Dirty Pictures
- Simulated annealing: theory and applications.
- Image segmentation techniques
- Markov Random Fields and Their Applications
- Simulated Annealing: Theory and Applications
- Probabilistic Solution of Ill-Posed Problems in Computational Vision
- Modeling and Segmentation of Noisy and Textured Images Using Gibbs Random Fields
- Simultaneous parameter estimation and segmentation of Gibbs random fields using simulated annealing