Abstract: The accurate extraction of optic disc contour is important to the diagnosis of eye disease.Nowadays,the research on optic disc contour extraction is particularly important because the eye diseases has taking serious injury to human.With the rapid development of information technology,it is quite common to use computer in medical image processing.With the help of the computer,this paper uses the Markov with Field (MRF) theory to model the fundus images,and combines with the K-means algorithm,condition iterative algorithm and the expectation maximization algorithm and so on to achieve the segmentation of optic disc.The it extractes the optic disc through the MPP algorithm.Experimental results show that this method has higher segmentation efficienay and better denoising performance compared with the traditional segmentation methods.