Automated Blood Group Detection Using Fingerprint Biometrics and Deep Learning
Blood group identification is an important requirement in transfusion medicine, emergency care, surgery, and healthcare record management. Conventional ABO and Rh typing is highly established and normally relies on a biological blood sample and serological reactions. This paper presents an artificial-intelligence-based research prototype that investigates whether fingerprint images can be used as a non-invasive input for predicting the eight common ABO/Rh classes: A+, A−, B+, B−, AB+, AB−, O+, and O−. The proposed pipeline accepts a fingerprint image, performs image quality enhancement and normalization, extracts discriminative ridge features using a convolutional neural network (CNN), and assigns the image to a blood-group class. The system is designed as a screening and decision-support prototype rather than a replacement for clinical blood typing. Recent studies have reported promising classification performance on fingerprint datasets, while other clinical and dermatoglyphic investigations have found inconsistent or statistically weak relationships between fingerprints and blood groups. Therefore, this work emphasizes reproducible image processing, supervised learning. KEYWORDS Fingerprint Analysis, Blood Group Prediction, Machine Learning, Image Processing, Deep Learning, Classification