Finger Knuckle Print Recognition
using MMDA with Fuzzy Vault
MuthuKumar Arunachalamand and Kavipriya Amuthan
Department of Electronics and
Communication Engineering, Kalasalingam Academy of Research Education, India
Abstract: Currently frequent biometric scientific research such
as with biometric applications like face, iris, voice, hand-based biometrics
traits like palm print and fingerprint technique are utilized for spotting out
the persons. These specific biometrics habits have their own improvement and
weakness so that no particular biometrics can adequately opt for all terms like
the accuracy and cost of all applications. In recent times, in addition, to
distinct with the hand-based biometrics technique, Finger Knuckle Print (FKP) has
been appealed to boom the attention among biometric researchers. The image template
pattern formation of FKP embraces the report that is suitable for spotting the
uniqueness of individuality. This FKP trait observes a person based on the
knuckle print and the framework in the outer finger surface. This FKP feature determines
the line anatomy and finger structures which are well established and persistent
throughout the life of an individual. In this paper, a novel method for
personal identification will be introduced, along with that data to be stored
in a secure way has also been proposed. The authentication process includes the
transformation of features using 2D Log Gabor filter and Eigen value
representation of Multi-Manifold Discriminant Analysis (MMDA) of FKP. Finally,
these features are grouped using k-means clustering for both identification and
verification process. This proposed system is initialized based on the FKP
framework without a template based on the fuzzy vault. The key idea of fuzzy vault
storing is utilized to safeguard the secret key in the existence of random
numbers as chaff pints.
Keywords: Finger Knuckle Print (FKP),2D Gabor filter,
Multi-Manifold Discriminant analysis (MMDA),Fuzzy Vault
Received January 5, 2018;
accepted December 17, 2019