Niyati Ajay Dave

• Fingerprints contain small details called minutiae, key features for comparing fingerprint patterns. This thesis develops a framework that extracts and matches minutiae points for fingerprint matching. It applies preprocessing steps to enhance detection accuracy across fingerprints of varying qualities. The YOLOv8 model detects key points, and the fusion of You Only Look Once v8 (YOLOv8) with Scale-Invariant Feature Transform (SIFT) generates descriptors for matching. Fast Library for Approximate Nearest Neighbors (FLANN) performs descriptor matching using Lowe’s ratio test, followed by Random sample consenus (RANSAC) to remove incorrect matches and refine the results. The framework test on diverse datasets, and the mean Average Precision (mAP@0.5) evaluates the performance of the YOLOv8 model in fingerprint-matching.