Published Paper Details:

Using machine learning models and fingerprint pattern analysis, this study introduces a novel biometric-based method for blood group prediction. The suggested system takes pictures of fingerprints, preprocesses them to improve the ridges, and then extracts features like texture parameters, orientation field, and ridge count. A supervised classification model that predicts the subject's Rh and ABO blood groups is trained using these features. The CNN-based model outperforms more conventional classifiers like KNN and SVM, according to experimental evaluation using a custom dataset. The findings show a strong relationship between blood group inheritance patterns and dermatoglyphic traits. By combining genetic trait analysis and image processing, this study advances the rapidly developing field of biomedical biometrics and opens the door to effective, non-invasive medical diagnostics. This study uses a hybrid image-processing and deep-learning pipeline to study the non-invasive prediction of ABO and Rh blood groups from fingerprint images. We examine dermatoglyphic correlations that have been documented in the literature, explain how to collect and preprocess datasets, extract handcrafted features (such as ridge density, minutiae statistics, and texture descriptors), and contrast end-to-end Convolutional Neural Networks (CNNs) with traditional machine-learning classifiers (SVM, Random Forest, and KNN). With an accuracy of approximately 82.3 Index Terms--Blood group detection, fingerprint analysis, machine learning, biometrics, biomedical engineering