Automated Blood Group Detection System Using Image Processing for Non-Invasive Antigen Feature Extraction and Classification

The project intends to design an RFID based automated blood group detection system through image processing and a specific CNN. First, the provided images of blood cells are processed and the system extracts features relevant to the given antigen for the purpose of identifying the blood group. With the help of a set of 12,500 augmented images of blood cells, the System processes and normalizes these images for classification. The presented method can be considered as more effective to detect blood groups in comparison with the traditional method that requires invasive procedures and requires manual inspection. It is a system that optimizes the detection of blood groups leading to improvements in medical diagnosis and medical imaging in general.