Optimal Fuzzy Clustering by Improved Biogeography-based Optimization for Leukocytes Segmentation

Automated leukocytes segmentation is a difficult process in case of histopathological images due to the presence of various morphological structures in the images. This paper proposes a new fuzzy clustering method using improved biogeography-based optimization method which segments the leukocytes in the histopathological images of inflamed rat skins. The proposed method uses the ratio of compactness and separation as an objective function for getting the better cluster quality. The performance analysis of the proposed clustering method is done on histopathological inflamed rat skin images using F1 score and aggregated jaccard index value. The experimental results depict that the proposed method outperforms the other existing clustering based segmentation methods.