CSDL Home IEEE Transactions on Pattern Analysis & Machine Intelligence 2011 vol.33 Issue No.01  January
Subscribe
Issue No.01  January (2011 vol.33)
pp: 1629
Frederic Lehmann , Institut TELECOM, TELECOM SudParis, Evry
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/TPAMI.2010.58
ABSTRACT
We consider the problem of semisupervised segmentation of textured images. Existing modelbased approaches model the intensity field of textured images as a GaussMarkov random field to take into account the local spatial dependencies between the pixels. Classical Bayesian segmentation consists of also modeling the label field as a Markov random field to ensure that neighboring pixels correspond to the same texture class with high probability. Wellknown relaxation techniques are available which find the optimal label field with respect to the maximum a posteriori or the maximum posterior mode criterion. But, these techniques are usually computationally intensive because they require a large number of iterations to converge. In this paper, we propose a new Bayesian framework by modeling twodimensional textured images as the concatenation of two onedimensional hidden Markov autoregressive models for the lines and the columns, respectively. A segmentation algorithm, which is similar to turbo decoding in the context of errorcorrecting codes, is obtained based on a factor graph approach. The proposed method estimates the unknown parameters using the ExpectationMaximization algorithm.
INDEX TERMS
Texture segmentation, Markov random field, hidden Markov autoregressive model, factor graph, forwardbackward algorithm, turbo processing.
CITATION
Frederic Lehmann, "Turbo Segmentation of Textured Images", IEEE Transactions on Pattern Analysis & Machine Intelligence, vol.33, no. 1, pp. 1629, January 2011, doi:10.1109/TPAMI.2010.58
REFERENCES
