Implementation of Blood Cell Counting Algorithm using Digital Image Processing Techniques

Blood is a connective tissue with cellular components like RBC, WBC, and platelets. If the blood cells are below the standard range it leads to varies health diseases and hence proper blood cell counting technique plays importance in the pathology department. The most widely used method for counting blood cells is the microscopic technique which gives better results however, it takes more time to count the number of blood cells. Blood cell count using a digital image processing technique can be considered to automate the microscopic method and minimize the time taken by the conventional method. In this paper, a digital image processing technique is proposed to accurately count all the blood cells in a given slide with maximum accuracy. A method such as morphological operator, texture-based classification method and Circular Hough Transformation (CHT) used to count the RBC which gives accuracy around 90%. Similarly, Morphological Operator, Vegetation method, and Texture Object-Based classification are implemented to count the WBC and Platelet. However, Texture Object-Based classification method gave 100 % accuracy compared to Morphological Operator and Vegetation method. From the results discussed in this paper conclude that for RBC count using Circular Hough Transformation gives better result and WBC, platelets count using the Texture Object-Based classification method gives better results.