An Effective Clustering Technique for WBC Image Segmentation and Its Classification

For proper diagnosis of the disease, the immature white blood cells(WBC) have to be detected. As the first process, segmentation has to be done to extract the nucleus of white blood cell image. The main task of the segmentation is to extract the meaningful objects from an image. During segmentation, the size and shape of the nucleus should be maintained. In image segmentation applications, clustering techniques are commonly used to segment regions of interest or detect border of objects in an image. Many segmentation algorithms have been developed for various applications. The modified versions of moving k-means algorithm called adaptive fuzzy moving k -means which is the combined technique of fuzzy moving k-means and adaptive moving k-means is employed here for segmentation. This algorithm result is compared with the conventional clustering algorithms. The main objective is to segment the WBC from the blood smear image to detect an immature cell.