Section Article

Fingerprint Recognition Using Principal Component Analysis and Various Distance Measures
Author(s): Dr. Vikas Sangwan

Abstract
Fingerprint recognition has expanded into one of the most widely deployed and technically sophisticated biometric identification methods used globally across security surveillance digital authentication border management and mobile device verification systems. Between 2000 and 2018 significant advances in both computational intelligence and image-processing methodologies enhanced the accuracy reliability and robustness of fingerprint recognition systems. At the heart of these improvements lies the integration of dimension-reduction algorithms—especially Principal Component Analysis (PCA)—and multiple distance-based classifiers that help in identifying the similarity between fingerprint features. PCA offers an efficient strategy to extract discriminative global features from high-dimensional fingerprint images reducing computational complexity while preserving critical variance information. When combined with various distance measures—such as Euclidean Manhattan Mahalanobis correlation-based cosine similarity and Minkowski metrics—the system transforms into a powerful recognition framework capable of producing high identification rates even under unfavorable imaging conditions.