Elevating Fingerprint Matching with a Unified dHash, SIFT, and LSH Model

User authentication is essential for maintaining the security, privacy, and integrity of systems and data. It ensures that only authorized individuals can access specific resources, applications, or sensitive information, preventing potential data breaches, unauthorized modifications, or malicious activities. Biometric identification, particularly fingerprint recognition, has revolutionized authentication. This research paper proposes combining cutting-edge algorithms such as dHash, SIFT-Harris feature points, and Locality-Sensitive hashes to create a more efficient fingerprint-matching system. It aims to improve matching accuracy, data breach avoidance, robustness, and processing speed in fingerprint recognition systems. The final proposed model works for almost all altered (obliteration, central rotation, and z-cut) images. A total of 55,273 images were tested, giving an accuracy of 0.99.