HemoDetect: Blood Group Detection from Fingerprint Using Machine Learning and Deep Learning Methodologies

Fingerprints are a unique identification mechanism that is useful for determining human identity. Hardly one in more than 64,000 million people share a similar fingerprint pattern. The minute pattern of the fingerprint of each individual is unmatched with minimal possibility of similarity. The science behind detecting the blood group from the fingerprint is derived from the presence of proteins and antigens present in sweat on the ridges and grooves of the fingerprint. This paper uses advanced image processing and machine learning algorithms to compare fingerprint patterns and various blood groups. This research aims to contribute to biometric applications and thus improve healthcare through innovative approaches using the latest technology, such as artificial intelligence. Deep learning methodologies such as the Convolutional Neural Network (CNN) extract hidden patterns inside the image and make the necessary predictions. Traditional blood group detection methods are time-consuming and can cause human errors. To make blood group determination error-free and less time-consuming, various machine learning and image processing techniques can be implemented. The detection of blood groups is important for cases such as medical diagnosis, blood transfusions, etc. This study shows the correlation between fingerprint patterns and the different blood groups.