Abstract
Unsupervised classification of the brain MRI image into different tissue regions shows salient role to analyze the neuro-imaging and scientific research in the field of medicine and diagnosis. In this paper, an efficient unsupervised spatial fuzzy c-means technique is proposed for classification of distorted brain MRI image with different bias factors. The spatial information has been measured by analyzing three different neighbouring distances to suppress the noise and inhomogeneity as well as preserving the image details at the time of classification. The correlation among the neighbours have been decided by spatial information so that if the centers are itself affected with bias error, it would not affect the correlation. The qualitative and quantitative results are used to state the superiority of the proposed method from the considered methods of interest with reference to the partition coefficient and partition entropy.
