Mixture model based image segmentation with spatial constraints

One of the many successful applications of Gaussian Mixture Models (GMMs) is in image segmentation, where spatially constrained mixture models have been used in conjuction with the Expectation-Maximization (EM) framework. In this paper, we propose a new methodology for the M-step of the EM algorithm that is based on a novel constrained optimization formulation. Numerical experiments using simulated and real images illustrate the superior performance of our methodology in terms of the attained maximum value of the objective function and segmentation accuracy compared to previous implementations of this approach.