Abstract:
Biometrics refers to the automatic recognition of individuals
based on their physiological and behavioural characteristics. These
characteristics are unique to each individual and remain unaltered
throughout human lifetime. Several unimodal, bimodal and trimodal
biometric security systems have been developed using convolutional
neural network but few of them have been able to handle the
challenges of accurate recognition rates and processing time. In this
work, a comparative study of the performances of unimodal, bimodal
and trimodal biometric security systems using deep learning
technique was carried out.
The System was tested on a database consisting of 1026
trained images and 684 probe images of face, ear and iris biometrics.
All the images were preprocessed. Feature extraction and
classification were carried out using deep learning technique,
precisely, Convolutional Neural Network Algorithm (CNN). The
results show that the unimodal, the Ear system produced highest
value of 91.67% accuracy, Sensitivity of 93.57%, Specificity of
85.96%, Precision of 95.24% in in 114.10secs time. In bimodal
system Ear-Iris produced highest value of 96.05 accuracy, Sensitivity
of 96.49%, Specificity of 94.74%, Precision of 98.21% in 297.01
time, while the developed system propduced Sensitivity of 97.66%,
Specificity of 98.25%, Precision of 99.40%, Recognition Accuracy of
97.81% but the Recognition Time of 455.54 Secs
Download full Lenght Paper......