In this paper we propose an image segmentation algorithm based on new-symmetric mixture model. Here the pixel intensities of the whole image are characterized through a newsymmetric mixture distribution, such that the statistical characteristics of the image coincide with that of the new symmetric distribution. Using the K-Means algorithm the number of image regions and initial estimates of the model parameters for the EM algorithm are obtained. The segmentation algorithm is proposed by component maximum likelihood under Bayesian frame work. The efficiency of the proposed method is studied with the five images taken from the Berkeley image dataset and computing the values image segmentation measures like global consistency error, probabilistic rand index and variation of information. A comparative study of the proposed model with Gaussian mixture model reveals that the proposed method performs better. The efficiency of the proposed method with respect to the image retrieval is also studied.