WHITE BLOOD CELL CLASSIFICATION AND COUNT USING CNN AND IMAGE PROCESSING

Sudarshan Dhakal
2020
BSc.CSIT
Semester 7
Downloads 2

White blood cells in our blood stream provides a glimpse into the state of our immune system and any potential risk we might be facing. White blood cells analysis is generally performed in evaluating hematic pathologies such as immune deficiency syndrome (AIDS), blood cancer (leukemia) and other related disease [1]. In particular, a dramatic change in white blood cells count relative to baseline is generally a sign that your body is currently being affected by an antigen [2]. Generally, a variation in a specific type of white blood cells correlates with a specific type of antigen; people having allergies generally see an increase in their eosinophil counts as they are responsible for fighting allergens [3]. Therefore, counting and classification of white blood cells in the most efficient way is very important. Unfortunately, the traditional manual method is very time consuming, inaccurate and tedious [4]. In traditional method, people manually classify and count the white bloods cells. Hence, system is needed to classify and count the different types of white blood cells. In this software, the blood cells are classified and counted using the microscopic image of white blood cells. In the classification and counting, the features of image is extracted using CNN and Image processing algorithms. Moreover, three different types of features were extracted which are morphological, statistical and textural. Therefore, the overall result confirms that the software can produce accurate result in less period of time; it can be applied in hematological laboratories. The average accuracy was 85%.

white blood cells
classification of wbc
leukocytes
blood cell count

Similar Projects