Optimizing Fingerprint Identification: CNNs With Raw Images Versus Handcrafted Features for Real-Time Systems

Fingerprint identification is a cornerstone in various domains such as security, forensics, and authentication. Despite progress, existing systems still face challenges with noise, database differences, and real-time speed. This study investigates the balance between accuracy and computational effic...

Full description

Bibliographic Details

Published in:	IEEE Access
Main Authors:	Shaik Salma, Tauheed Ahmed, Garimella Ramamurthy
Format:	Article
Language:	English
Published:	IEEE 2025-01-01
Subjects:	Convolutional neural network handcrafted feature vectors genuine matching impostor matching minutiae-based features
Online Access:	https://ieeexplore.ieee.org/document/11037729/