Published Paper Details:

The blood cell classification from smear images of peripheral also known as PBS is a crucial step in diagnosing blood-related illnesses such as anemia, leukemia, infection, polycythemia and malignancy. Hematologists regularly utilize a device which zooms to microscopic level to count, shape, and distribute the cells before making a judgment in blood cell-based analysis. Hematology analyzers and flow cytometry give an accurate and precise CBC that identifies and gives the abnormalities in given smear slides of blood. In multiple hospitals, the techniques that are used are costly, least effective in terms of time and are hectic as they are manual. As a result, a reliable, affordable, and automatic method for identifying different sickness through given PBS image is required. The new proposed model automatically does the examination of the provided data, also does in a faster manner. Therefore, in the studied research we have properly and accurately done the classification of images in such a way that the different cells are classified as the WBCs and RBCs. Feature extraction is done to classify the images. These extracted texture features are subsequently inputted into various classifiers, including artificial neural networks (ANN), machines that are SVM and many other model which are in ML and DL. After comparing the performance metrics, it is determined that the logistic regression method is the most appropriate for the task at hand.