A simple and accurate method for white blood cells segmentation using K-means algorithm
Segmentations of leukocyte play an interesting research topic in medical image processing especially in Hematology and related study field. In this paper, we present a segmentation of nucleus and cytoplasm in adult white blood cell from RGB color images. The cell segmentation begins by applying a dilated perimeter of nucleus convex hull which extend into a surrounding region in order to setup a color reference table of cytoplasm. Primary cytoplasm region is then estimated approximately. Distance mapping is applied to this primary area and converted into a gradient vector flow. The active contouring technique is then implemented respect to the vector field and finally terminate at the cell edge. The obtained segmentation results show that active contour which guided by distance mapping from a neighboring area are able to extract nucleus and cytoplasm region promisingly.