Automatic Evaluations of Human Blood Using Deep Learning Concepts
Geeta Rani, Harini Mohan, Bendela Kusuma, Pranay Kumar, Ardhala Mounika Jenny, Nukala Akshith
Abstract
Identification of human blood is very important to know before any blood transfusion. In emergency situations, if the blood group is not identified exactly, it causes many problems to the patient and may become fatal. Generally, the blood group is identified manually by the lab technicians performing blood tests. In all times, humans cannot be perfect, and in this pandemic, lab technicians need to handle large number of blood samples. So, the manual identification may undergo human errors. The idea of the present paper tends to solve the human errors using deep learning techniques. The proposed system identifies the blood group of each person and reduces the human errors. This would save the time consuming for testing and gives efficient results with good precision and accuracy. Deep learning obtains the blood sample from blood donation applications and have a trained model to predict the given image. Using the latest image processing models, the blood samples are detected. The result through the proposed system is a set of clustered blood samples. Thus, the proposed system benefits the society and efficiently affects the medical diagnosis.