Programmed Blood Boundary Identification Utilizing Matlab

Abstract

Blood and its components are main indicator tool in determining many pathological conditions. The deficiency of red blood cells, which constitutes 99% of blood cells and specialized as an oxygen carrier, causes various blood disorders. The diagnosis of blood cells manually is tedious and time consuming that could be simplified utilizing automatic analysis. We aim to automate the process by means of making a system that will detect RBC and WBC count along with detecting the level of hemoglobin and cancer cell detection. We aim to use processing along with machine learning to create this system that will provide RBC as well as WBC count of the user and will provide an accurate hemoglobin level for the sample provided by the user. We will use multiple image processing techniques and machine learning technique to achieve this desired result. With the advent of image processing, neural network microscopic photographs can be processed for the parameters required. The input of the system will be an image of the blood sample that will be acquired using microscopic camera and will be stored in database. Various algorithm are being applied on the image to count different parameters from it and accordingly the output will be displayed on the screen. In simple words, this system will prove to be cost effective and will not be time consuming as it will be using the best algorithms to automate the system.