Image Segmentation Using a Hybrid Clustering Technique and Mean Shift for Automated Detection Acute Leukaemia Blood Cells Images

Clustering is one of the most common automated segmentation techniques used in the fields of bioinformatics applications specifically for the microscopic image processing usage. Recently many scientists have performed tremendous research in helping the haematologists in the issue of segmenting the leukocytes region from the blood cells microscopic images in the early of prognosis. During the post processing, image filtering can cause some discrepancies on the processed image which may lead to insignificant result. This research aims to segment the blood cell microscopic images of patients suffering from acute leukaemia. In this research we are using three clustering techniques which are (Fuzzy C-Means (FCM), Classic K-Means (CKM) and Enhanced K-Means (EKM) then we performed filtering techniques which are Mean-shift Filtering (MSF) and Seeded Region Growing (SRG). We tested individual clustering, from the results it show Enhanced K-Means gives the best result. We performed hybrid between EKM and MSF gave a better result from other comparison. The integrated clustering techniques have produced tremendous output images with minimal filtering process to remove the background scene.