A Statical Color Image Segmentation Using a Diagonal Of The Modified Riesz Mixture Model

This paper describes a new approach to adapted estimation of parametric mixture model based on diagonal of the modified Riesz distribution (DMRD) defined in R , r ≥ 2. The DMRD can model accurately a withe variety of color image. This parameters index a family of distribution witch include the bivariate Gamma and the convolution product between bivariate Gamma and univariate Gamma. In our work, we have applied the bivariate Gamma distribution to mixture model. We use the Expectation Maximization (EM) algorithm to estimate the model parameters of the color image data and the number of mixture components is estimated by using K-means clustering algorithm. The K-means clustering algorithm is also utilized for developing the initial estimates of the EM-algorithm. This study investigates the DMRD in unsupervised color image segmentation. Experiments with real color images are described which verify the validity of method and its implementation.