Robust Image Segmentation with Mixtures of Student's t-Distributions

• Published in: 2007 IEEE International Conference on Image Processing Vol. 1; pp. I - 273 - I - 276
• Main Authors: Sfikas, G., Nikou, C., Galatsanos, N.
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.09.2007
• Subjects: clustering; Clustering algorithms; Coherence; Computer science; EM algorithm; Image segmentation; Inference algorithms; Maximum likelihood estimation; mixture model; Pixel; Probability density function; Robustness; segmentation evaluation; Student's t-distribution; Tail
• ISBN: 9781424414369; 1424414369
• ISSN: 1522-4880
• DOI: 10.1109/ICIP.2007.4378944

Gaussian mixture models have been widely used in image segmentation. However, such models are sensitive to outliers. In this paper, we consider a robust model for image segmentation based on mixtures of Student's t -distributions which have heavier tails than Gaussian and thus are not sensitive to outliers. The t -distribution is one of the few heavy tailed probability density functions (pdf) closely related to the Gaussian, that gives tractable maximum likelihood inference via the Expectation-Maximization (EM) algorithm. Numerical experiments that demonstrate the properties of the proposed model for image segmentation are presented.