BLOOD CELL SEGMENTATION IMAGES FROM MICROSCOPIC BLOOD IMAGES

Abstract

Blood cell segmentation is an important innovation for automatic differential blood counting, classification and analysis in clinical examination. Segmentation and recognition in colored blood cell images are two important issues in the field of biomedical cell morphology. This article approaches methods for distinguishing blood cells from microscopically thin blood images. This information is the basis for high-level tasks, such as automatic differential blood count, diagnosis of various diseases such as malaria, babesia, Chagas disease, anemia, leukemia. A system based on cell segmentation is an effective way to segment different blood components such as red blood cells (erythrocytes), white blood cells (wbc). Identification of parasites present in platelets as well as blood cells from images entered for blood count. Input images are captured by connecting a digital camera to a microscope. The resulting images are expanded and segmented using k-means clustering as well as global thresholding. Overlapping cells are separated using sobel edge detector and watershed transform.