Liver Segmentation using Weighted Contrast based Chan-Vese Method

Liver cancer is the leading causes for death globally. Segmentation of liver from liver CT images is very crucial to evaluate the diagnostic pattern of liver disease and is claimed to be a challenging task owing to the complexity of the liver shapes in different slices of CT. In this work, Chan vese level set segmentation algorithm coupled with contrast driven elastic optimization model is used to achieve good segmentation accuracy. Global contrast approach has been implemented in bottom-up saliency detection in curvature optimization. Saliency map and weighted coefficient technique is measured using mean shift filter which improves the accuracy of saliency detection. The proposed segmentation technique has accurate results compared to the existing segmentation technique.