Biometric systems are widely used in various applications of
today’s authentication technology. The unimodal systems suffer from var-
ious stumbling blocks such as noisy inputs, non-universality, intra-class
variability and imposter spoofing which affects the system performance
and accuracy. To effectively handle these problems, two or more individual
modalities are used. In this paper, we presented a multimodal approach
for fingerprint verification based on a combination of score level fusion
rules. In the preprocessing stage, Anisotropic Diffusion Filter (ADF) and
Histogram Equalization (Hist-Eq) techniques were applied to overcome
the main challenging drawbacks of fingerprint samples acquisition such as
distortion, noise, rotation, etc. Supplementary, the Local Binary Pattern
(LBP) was used for feature extraction. In score level fusion, the matching
scores of individual fingerprints were combined via several fusion rules.
Receiver Operating Characteristics (ROC) curves were formed for the
multimodal approach that’s why it is mainly used to evaluate our system.
Experimental results shown improvements of the multimodal system using
ADF and Hist-Eq versus the unimodal non-preprocessed fingerprint sam-
ples. The obtained results indicated that there is a significant increase in
the performance of the proposed system due to the combination of scores,
making it suitable for more applications relevant to identity verification.