Xiaogen Zhou, Zuoyong Li, Huosheng Xie*, Ting Feng, Yan Lu, Chuansheng Wang and Rongyan Chen
Leukocyte (White Blood Cell, WBC) segmentation, which is significant for white blood cell count and analysis, is a challenging task due to the morphological diversity of WBCs, the complex background, and different staining conditions. This paper presents a novel leukocyte segmentation method based on adaptive histogram thresholding (AHT) and contour detection. The proposed method first localizes leukocytes by image component combination and image thresholding, and crops sub-images located by leukocytes using a centroid-based method. Then, the proposed method employs AHT to extract leukocyte nucleus, and utilizes image color features to remove the complex backgrounds such as red blood cells (RBCs) and substantial dyeing impurities. Finally, Canny edge detection is performed to extract the entire leukocyte. Accordingly, cytoplasm segmentation is achieved by subtracting the WBC nucleus region from the leukocyte region. Experimental results on the dataset containing 160 leukocyte images show that the proposed method achieved more accurate segmentation results than the counterparts.
leukocyte (WBC) segmentation, leukocyte localization, image component combination, adaptive histogram thresholding, Canny edge detection, morphological operation.
College of Mathematics and Computer Science, Fuzhou University, Fuzhou, P.R., Fujian Provincial Key Laboratory of Information Processing and Intelligent Control, Minjiang University, Fuzhou, P.R., College of Mathematics and Computer Science, Fuzhou University, Fuzhou, P.R., College of Mathematics and Computer Science, Fuzhou University, Fuzhou, P.R., College of Mathematics and Computer Science, Fuzhou University, Fuzhou, P.R., School of Computer Science and Technology, Harbin University of Science and Technology, Harbin, P.R., Department of Clinical Laboratory, the People`s Hospital Affiliated to Fujian University of Traditional Chinese Medicine, Fuzhou, P.R.