Skin color segmentation plays a vital role in different applications such as Face Detection,Face Recognition and Human Computer Interaction applications. To improve the accuracy of skin colour segmentation system,in this paper a novel and new skin colour segmentation algorithm is proposed based on statistical approach under HSI colour space of the image. The bivariate feature vector of the image is to be model with a Pearson type II αα mixture (bivariate Beta mixture) model. The model parameters are estimated using EM Algorithm. The initialization of parameters is done through Hierarchical Clustering and moment method of estimation. The performance of the proposed skin colour segmentation algorithm is studied by computing the image segmentation quality metrics (PRI,VOI and GCE) and comparing them with that of bivariate Gaussian mixture model. The ROC curves plotted for the system also revealed that the proposed algorithm can segment the skin colour more effectively than the existing segmentation algorithm for classifying skin colour. © 2016,International Journal of Pharmacy and Technology. All rights reserved.