Color Image Segmentation Using a Semi-wrapped Gaussian Mixture Model
This article deals with color image segmentation in the hue-saturation-value space. Hue, saturation and value components are samples on a cylinder. A model for such data is provided by the semi-wrapped Gaussian distribution. Further its mixture is used to approximate the hue-saturation-value distribution. The mixture parameters are estimated using the standard EM algorithm. The results are obtained on Berkeley segmentation dataset. Comparisons are made with vM-Gauss mixture model, GMM and Mean-Shift procedures. Experimental results reveal improvement in segmentation by our method.
KeywordsMixture Model Gaussian Mixture Model Multivariate Gaussian Distribution Mixture Parameter Color Image Segmentation
- 1.Tao, W., Jin, H., Zhang, Y.: Color image segmentation based on mean shift and normalized cuts. IEEE Trans. on SMC(B) 37, 1382–1389 (2007)Google Scholar
- 6.Roy, A., Parui, S.K., Roy, U.: A color based image segmentation and its application to text segmentation. In: Proc. of Indian Conf. on Computer Vision, Graphics and Image Processing, pp. 313–319 (2008)Google Scholar
- 7.Martin, D., Fowlkes, C., Tal, D., Malik, J.: A database of human segmented natural images and its application to evaluating segmentation algorithms and measuring ecological statistics. In: Proc. Int Conf. Computer Vision, pp. 416–423 (2001)Google Scholar
- 8.Chabrier, S., Laurent, H., Rosenburger, C., Marche, P.: A comparative study of supervised evaluation criteria for image segmentation. In: Proc. of the European Signal Processing Conference, pp. 1143–1146 (2004)Google Scholar