International Journal of Scientific Research and Engineering Development

Determining an individual’s blood type is an important procedure in medical practice, especially in emergencies where quick and precise results can save lives. Conventional methods usually involve invasive lab-based examinations, which are usually time-consuming and can slow down urgent medical decisions. This project introduces a non-invasive, a completely automated method for identifying blood groups using fingerprint images.The system relies on a Convolutional Neural Network (CNN), which processes fingerprint images to deliver quick and reliable predictions. Specifically a customized ResNet9 model, which enables accurate and rapid prediction,to process fingerprint images and classify them into the four main blood groups: A, B, AB, and O. A Flask-based backend is integrated with a web interface, enabling users to upload fingerprint images and instantly receive results