Segmentation of color image using deterministic annealing EM

We present a color image segmentation algorithm based on statistical models. A novel deterministic annealing EM (expectation maximisation) formula is used to estimate the parameters of the Gaussian mixture model which represents the muti-colored objects statistically. The experimental results show that the proposed deterministic annealing EM provides a global optimal solution for the maximum likelihood parameter estimation and the image field is segmented efficiently by using the estimates.