Automatic detection of leukocytes for cytometry with color decomposition

Abstract The automatic detection of leukocytes is essential for cytometry and type recognition. Although this problem has been given much attention, the accuracy and speed of leukocyte detection still needs to be improved to meet the requirements of practical applications. This paper presents an effective method of detecting leukocytes. In this method, color decomposition and adaptive binarization are firstly used to detect the background and red blood cells, then the leukocytes can be roughly detected. After denoising, border refinement is applied to modify the false border of the roughly detected leukocytes, and a novel nucleus-enhancing method is used to identify nucleus, which can be used to detect false leukocytes. The proposed leukocyte detection method is evaluated using the blood cell image data set. The experimental results demonstrate that the proposed method outperforms two traditional methods, namely the region growing and snake methods, with higher accuracy and shorter computing time.