Tumors in liver have become a disease of major concern in recent times. Detection of the tumor and finding whether it is cancerous or non-cancerous is a crucial task for diagnosis of liver tumor as there is no established framework for evalution of medical imaging segmentation till now. Segmentation techniques can help in detecting these and also separating the region of interest from the rest of the neighbouring tissues of same intensity. This paper presents a method for detecting tumors in liver and separating the liver organ from its neighboring organs using Marker Controlled watershed segmentation. It also presents a classification model to categorize the tissues into tumored and untumored cells. This technique could correctly detect and classify the tumor tissues in liver in the patients’ computed tomography images. The accuracy of the classification model is satisfactory and the estimated sensitivity and specificity correctly validates the segmentation and the classification model used for this research