live ness, Real, Fake, Logical
Regression,Fingerprint validation.
Abstract
Fingerprint based recognition systems have been widely deployed in numerous civilian and government applications. However, fingerprint based biometric systems are vulnerable to spoofing attacks. To protect against spoofing, methods of liveness detection measure physiological signs of life from finger prints, ensuring that only live fingers are captured for enrollment or authentication. In this project, propose a novel image descriptor for fingerprint live-ness detect -ion using the local coherence of a given image. The key idea of the proposed method is that the difference between live and fake fingerprints is well revealed by the directional coherence in the gradient field. More specifically, the replica fabrication process is highly likely to yield dispersion of gradient distributions in a fake fingerprint image due to loss of information by the acquisition noise and blurring artifacts. Therefore, propose to exploit the local patterns of the directional coherence as our features, the so-called local coherence pattern (LCP). As compared to previous approaches, the proposed method efficiently describes the Under -lying structure of ridges and valleys (i .e., directional structure) in the fingerprint image and thus successfully captures the slight difference of textural characteristics between live and fake fingerprints even under noisy environment The experimental results on various datasets demonstrate that the proposed method efficiently improves the performance of fingerprint live-ness detection.
Article Details
Unique Paper ID: 157740
Publication Volume & Issue: Volume 9, Issue 8
Page(s): 902 - 906
Article Preview & Download
Share This Article
Join our RMS
Conference Alert
NCSEM 2024
National Conference on Sustainable Engineering and Management - 2024