The process of image segmentation is always affected by noises present in the images. A segmentation method that segments the images even in the presence of noise is proposed. The segmentation process is initialized with the help of the fuzzy partioning of the images. Then clustering is employed to differentiate the image pixels. For clustering FCM is employed. The EM and MRF were applied to the clustered image to segment the image even in the presence of noise. The performance of the process is measured with the help of accuracy and error rate. Image segmentation is the process of obtaining particular regions from the images. Edge detection, Contour extraction, Clustering are all the most common segmentation methodologies. Edge detection identifies the edge points around the needed objects. Contour extraction refers to outlining the segmented portion from the image. The clustering based methodologies segments the regions by grouping the pixels that have similarities in any of the cases such as intensity, color etc. The process of image segmentation is always affected by noises present in the images. A segmentation method that segments the images even in the presence of noise is proposed. The segmentation process is initialized with the help of the fuzzy partioning of the images. Then clustering is employed to differentiate the image pixels. For clustering FCM is employed. The EM and MRF were applied to the clustered image to segment the image even in the presence of noise. The performance of the process is measured with the help of accuracy and error rate. Noises were added to the input image. The noisy input image is initially undergone Fuzzy partioning. The Fuzzy partioning identifies the different intensity regions in the images. The intensity differences were identified and then clustering process is employed for the separation of the different clusters. The clustering process is employed using Fuzzy c means clustering process. The clustered regions were then segmented using the Gaussian mixture model. The likelihood measured in the clustered results were maximized with the help of the EM algorithm. In EM algorithm the Gaussian distribution at each pixel ranges were measured. The Markov Random Model is employed using the identified maximum likelihood estimated using EM model which gives the maximized likelihood informations. The Gaussian mixture model is employed with the likelihood informations obtained using the EM model to cluster the input image into clusters. The performance of the process is finally measured by measuring the accuracy and error rate of the segmentation process.