Image Segmentation Quality Assessment For "Truncated Compound Normal With Gamma Mixture Model"

In this paper, we present image segmentation quality assessment study for  Truncated Compound Normal with Gamma Mixture(TCNGM) under Expectation Maximization(EM) framework. Segmentation quality metrics such as Global Consistency Error(GCE), Probabilistic Rand Index(PRI), and Variation of Information(VoI) are applied to the clusters of image pixel labels produced by the model in comparison to the other models used for image segmentation. We show that our model is a competing one with other models.