Histological Image Segmentation using Fuzzy C-Means

This paper deals with the automatic segmentation of Haematoxylin and Eosin(H&E)stained Histological slide image with the help of advanced soft clustering mechanism. The clustering mechanism used in this proposed framework is Fuzzy C-Means (FCM) algorithm and it is implemented on the human skin dataset. The dataset is obtained by digitally scanning the H&E stained histological slide of human skin tissue with the help of WSI (Whole Slide Image) scanner. The FCM clustering mechanism is implemented on the image obtained after converting to L*a*b* colour space. This paper presents a detailed discussion regarding the L*a*b* colour space followed by the soft FCM algorithm. The experiment is carried out on human skin tissue and the results obtained after segmentation is shown which will prove to be helping hands for the medical practitioners for identifying and extracting the Region of Interest (ROI) for the purpose of diagnosis.