LAUSR

Image segmentation is the significant tasks in maximum medical diagnosis tools. In recent years, perfect segmentation of medical images is of the great challenging issue. Maximum investigation in most of the areas exhibit that numerous individuals having brain tumors expired because of authentic circumstance for inaccurate recognition. The traditional K-Means algorithm suffers from numerous issue which results in less accurate segmentation, therefore an efficient approach need to be introduced. In this regard, the proposed methodology focused on the MRI medical images for segmentation. Dissimilar to the existing approaches, in this proposed approach a novel K-Means algorithm is introduced by means of incorporating the utmost dominant gray level of the image. This paper addressed the issue of arbitrarily selecting k no of pixels using dominant gray level of the image. The experimental consequences of the proposed approach showed that it has better performance in terms of the images when compared to the traditional K-Means approach.