05-10-2012, 01:35 PM
Multimodal BIOMETRIcs
MULTIMODAL BIOMETRICS - Copy.pptx (Size: 1.74 MB / Downloads: 44)
What is multimodal Biometrics?
Multimodal biometrics is the form of Biometrics
which uses more than one form of Biometric Identifiers.
The common biometric identifiers include:
Voice
Face Recognition
Hand Scanner
Finger Print
Iris and Retina Scan
Keystroke Dynamics
Signature
WHY MULTIMODAL biometrics?
Most biometric systems in real applications are unimodal which means they rely only on one form of identification. Eg. Voice, Finger Print, Face etc.
Unimodal systems are quite vulnerable to problems such as noisy data, non-universality and spoofing.
High False Acceptance Rate (FAR) and False Rejection Rate (FRR).
Limited Discrimination Capability
Lack of Permenance
Benefits of multimodal Biometrics
Increased level of security (Ultra Security)
More than Average Accuracy
Very Reliable because of using many independent biometric identifiers together
Deter Spoofing
Administrator can decide the level of security needed for the firm
Overcomes the problems of worn and unrecognizable data when using single identifier
COMPONENTS OF A MULTIMODAL BIOMETRIC SYSTEM
The Sensor Module – which captures the trait in the form of raw biometric data
The Feature Extraction Module – which processes the data to extract a feature set
The Matching Module – which employs a classifier to compare the extracted feature set with the templates in the database to generate matching scores
The Decision Module – which uses the matching scores to either determine an identity or validate a claimed identity
Various levels of fusion
Fusion at the Data or Feature Level – Feature sets originating from multiple sources/sensors are fused
Fusion at the Match Score Level - The scores generated by multiple classifiers pertaining to different modalities are combined
Fusion at the decision level: The final out put1 of multiple classifiers are consolidated via techniques such as majority voting