The paper proposes three different methods (M1, M2, M3) for generating cancelable biometric templates using visual secret sharing.
M1 uses one original biometric image as the secret and n-1 randomly chosen gray images as cover images. M2 uses one original biometric image as the secret and n-1 randomly permuted versions of the original image as cover images. M3 uses both the secret image and the cover images as randomly permuted versions of the original biometric image.
The proposed methods involve the following steps:
The performance of the proposed methods is evaluated using various metrics like correlation coefficient, mean square error, mean absolute error, PSNR, SSIM, NPCR, and UACI. The results show that M3 generates the best quality cancelable templates, with the secret shares not revealing any information about the original biometric.
To Another Language
from source content
arxiv.org
Key Insights Distilled From
by Manisha,Niti... at arxiv.org 04-25-2024
https://arxiv.org/pdf/2404.15394.pdfDeeper Inquiries