An Effective Medical Image Segmentation Using Random Walker Algorithm with Fuzzy Logic

Image segmentation has often been defined as the problem of localizing regions of an image relative to content (e.g., image homogeneity). A good segmentation requires being fast and accurate. A vast research work has been done in the area of segmentation which has given a lot of methods for segmentation each having different features for different types of images. Random walker method is an algorithm which segments the images based on user defined seeds and the probabilities of unseeded pixels. Nevertheless, alike many other segmentation methods, it can be too slow for real-time applications. In this paper we present a new method in which fuzzy logic is combined with fuzzy logic to make Random Walker Method fast and more accurate. We have done this experiment on different medical images which are considered as complex images and also, require more accuracy and clarity in texture.