Description:
SEMMS stands for Spatial EM for Markovian Segmentation.
SEMMS is an algorithm for

• Model-based unsupervised image segmentation, including the following models : Hidden Markov Random Field and mixture model
• Model selection for the HMRF model
• Simulation of a realisation of a Potts model
• Addition of independent Gaussian noise to a given classification

Load:

• semms source code
• Readme and image example

Comments:
This program is made available to the research community in order to make easier comparisons between our algorithm and others. Thus, you can use it free of charge. The only guidelines that we would ask you to follow are that :

• you do not make a commercial use of the program ;
• you would mention the program's source if you use it for a published paper.

Related links:

• Nathalie Peyrard's web page
• Florence Forbes' web page
• EM Procedures Using Mean Field-Like Approximations for Markov Model-Based Image Segmentation INRIA Technical Report
• SEMMS code has been built from the NEM code of Mo Van Dang

Contact: Nathalie Peyrard, Florence Forbes

(Last mod: january 31, 2005.)