A Comparative Study of White Blood cells Segmentation using Otsu Threshold and Watershed Transformation
Nancy M. Salem, Noorhan M Sobhy, Mohamed El Dosoky · Journal of Biomedical Engineering and Medical Imaging · 2016
The aim of white blood cells (WBC) segmentation is to separate leukocytes from other different components in the blood peripheral image. In this paper, a method to segment white blood cells from microscopic images is proposed. The proposed method consists of three stages; Pre-processing, segmentation, and finally post-processing. In the pre-processing step; the color correction is used to enhance the image. In the segmentation step; two techniques have been used which are Otsu threshold and watershed marker-controlled followed by feature extraction. Shape features are used to differentiate between single and grouped cells. Artifacts are removed in the post-processing step. Experimental results show that the accuracy is 99.3% and 93.3% for the watershed based and the Otsu threshold based methods respectively. Experiments demonstrate that watershed marker-controlled outperforms Otsu threshold in the segmentation of WBC.