Abstract:
Biometric modalities provides us the study of characteristics related to physiological
traits for the verification or identification of a person. Unimodal biometric arises
and disadvantages of security, accuracy, performance
and data handing. Using multimodal modalities, security
many shortcomings
imperfection, storage space
will be much higher but accuracy, performance perfection, storage space and rate
data handing are still a subject of research and discussion. In our current research
study we are working with multimodal biometric modalities of face and thumb
for the validation of any individual to obtain high security impression recognition
with modalities offace and thumb impression. System will better accuracy using two
be able to train face and thumb impression of persons, and for testing and validation
both modalities are necessary to input or define. We are using Discrete Cosine
Transform algorithm which collects features point of the sample, furthermore the
data will be classified by the Support Vector Machine Technique to give the correct
collected dataset. There outcome. Another contribution of our project is our own
lack of datasets which can provide us both facial images as well as thumb
were
impression of the same person. The results have better accuracy rate and two
modalities necessary to input for validation which (face and thumb impression) are
makes the system more secure an applicable in many desired cases.
