An Adaptive Color Image Segmentation Algorithm Based on Gaussian Mixture Model Applied to Mobile Terminal

This paper presents an adaptive color image segmentation algorithm based on Gaussian mixture model. Image edge posterior probability density is estimated with Gaussian mixture model. To derive image edge, we use variational approximation method to estimate Gaussian mixture model parameters. Finally, we give some image segmentation experiments to verify performance of our algorithm combined with priori position information. Experimental results show that our algorithm can be applied to color image segmentation on mobile terminal.