MR Image Segmentation Based on Fuzzy Markov Random Field
Imaging modality provides detailed information about anatomy. It is also helpful in diagnosing disease and
helps in progressive treatment. We present suitable algorithms for thresholding, fuzzy k-means, fuzzy c-means and fuzzy
Markov random field (MRF) that could improve results. Since the k-means and c-means aims to minimize the sum of
squared distances from all the points to their cluster centre, it should results in compact clusters. For this purpose we use
intra-cluster distance measure which calculates the median distance between a point and its clustered centre. The proposed
algorithms are evaluated and compared with fuzzy k-means, c- means and MRF methods both qualitatively and
quantitatively to improve their efficiency. From the proposed work, it is obvious that, these evaluations are not easy to
specify in the absence of any prior knowledge about resulted clusters.
Keywords— Segmentation, Thresholding, Fuzzy-k means,(FKM) Fuzzy-c means(FCM), Markov random field(MRF).