In the present, the development of medical imaging has a bottleneck. However, there are a lot of continuous images produced from the computed tomography (CT) equipment, and it will cost much time for the physician to check and analyze. In this paper development of this system by the lung, we proposed the efficient framework of texture acquisition and contour segmentation by some useful image processing methods. After more than two patient of experiment, the end results show that the performance of the method is feasible and accurate.