Article Preview
Top1. Introduction
In this modern and hi-tech era, there is a massive demand for methods to identify and authenticate genuine individuals. Till now, we are dependent upon Passwords and pin numbers for identification for ATM/Credit cards/Debit Cards/Utilities on the Internet. Remembering all the passwords is not an easy task, and these can often be forgotten, lost, or stolen.It is the demand of time that such technologies need to be upgraded for security and identification of an individual since the current methods are hard to remember, easily transferable and stolen, and have numerous weaknesses. Due to the shortcomings of these methods, biometrics came into consideration.
We can classify the biometrics system according to the number of features used:
In unimodal biometrics systems, only one characteristic(like face, gait, fingerprint, palm, or iris) is sufficient for the identification and recognition of an individual. But it has its limitations.
In multimodal biometrics systems, more than one feature is used for the identification of an individual, which makes these systems more effective due to the presence of multiple, independent characteristics. They are also able to meet the high-performance requirements imposed by various applications. These biometric systems consign the complication of non-singularity since different features ensure that a sufficient population can be accommodated in the dataset.
Also, they provide anti-burglary procedures by creating it hard for an interloper to spoof the multiple biometric traits of a legal user simultaneously. By checking the subject for a random set of biometric peculiarities, the system ensures that a“live” user is indeed present at the point of data gathering. Thus, a challenge-response type of authentication can be facilitated using multimodal biometric systems.
In this digital era, itis very easy to steal a person's identity. There are several types of biometric systems that require a person’s contact with sensors or devices. Numerous Biometric Recognition Systems are available in the market, which aims to recognize a person’s identity. A contactless biometric recognition system has proven to be more advanced with enhanced utility in this era. But the contactless biometric system has a comparatively lesser recognition rate than a contact one. The recognition rate is a significant issue, which should not be compromised, as compared to the ease of applicability (Frischholz, R. W. and Dieckmann, U.; 2000).
In this study, we developed a recognition system that aims to combine three different modalities, which improve the modality recognition rate when compared with Face, Ear, and Gait.individual modality recognition rate. All the single feature-based techniques have comparatively less recognition rate; however, the combination of these techniques gives impressive results so that they can be implemented in legacy systems.
The computed results prove that the multimodal biometric system has a comparatively higher recognition rate than that of unimodal systems.
The objective of this study was to design a multimodal biometrics system using face, ear and gait biometrics, to finally recognize the person as genuine or an impostor. At the same time, we have to overcome the limitations of the unimodal biometric system and generate further reliable and robust system with multiple, independent traits.
This biometric multimodal approach is practically implemented by the Government of India for their citizen identification program called as AADHAR card where more than one biometric is used for the identification. The same can be implemented for verifications in crimes(murder,kidnapping,chain snatching,etc.) performed by the criminals.Securtiy agencies may initially create the database of criminals by taking the relevant information of multiple features (face, ear and gait) which can be verified later on to catch actual culprit. This database can be fetched anywhere across the wold if available on internet. If the offender even leaves the country, he/she can be catched with the help of this database because of these multiple stored features information.