Biometrics and Strong Authentication - Research Proposal Example

Biometrics Introduction The world is full of many people with different identities. Sometimes, people move from one place to another, and they engage in one activity or the other. There is always the need to identify and authenticate the identity of “who someone is” (Woodward 3). Biometrics is such a useful technique that makes it possible to authenticate someone’s identity. Biometrics largely depends on any automatically measured physical characteristics or personal traits that are peculiar to an individual (Woodward 7). These physical characteristics and personal traits include fingerprint, hand, face and voice recognition. But this research would be concentrating on the use of face and fingerprint to identify people. Methodology This report employs quantitative research technique by critically analyzing previously recorded or published works on the theme of Biometrics. The essence of this is to analyze, summarize and eventually draw a conclusion on the extent of usefulness of biometrics as a significant instrument for personal identification and authentication. Relevant literature is used to produce cogent reasoning for the importance of Biometrics in the modern world. Analysis Practical biometrics employs the use of both facial recognition and fingerprinting to authenticate a person’s identity. This unique system has been proved to be the most efficient method for personal identification. Fingerprinting has become the oldest technique used as biometric markers. These markers are “ the impressions of the papillary or friction ridges on the surfaces of the hand. In police and civil applications, the primary interest is in the ridges on the front of the fingers above the end joint. In certain forensic applications, the area of interest is broader and includes all of the friction ridge surfaces on the hands. This includes the fingers, the palms, and the so-called writers’ palms (that is, the area on the little finger side of each hand, the part that rests on the paper when one writes),” (Ware 75). However, the latent impressions that stuck on objects someone has touched or grabbed came from a residue deposited on the objects; and this residue is from a collection of sweat, organic solids such as amino acids, and inorganic solids such as salts or blood, or other materials the hand has touched lately. The earlier use of fingerprint card was the first attempt to standardized the fingerprint data gathered from individuals. FBI developed several standards to differentiate which fingerprint belongs to whom: these include the card sizes, type of ink, text fields, location of fingerprints, and other factors (Ware 78). Later, manual matching of fingerprints was used as law enforcement had to struggle to identify and authenticate many fingerprints collected. By 1961, there have been several publications on the automated comparison of fingerprints. And FBI worked hard to apply the use of automation to organize its fingerprinting data. During this period, M40 algorithm was developed: this was the first matching algorithm used operationally at FBI (Ware 81). French and British scientists, and the others conducted some researches to discover the best automation technique: And some high-performing equipment that could be used to scan, print, compare and differentiate fingerprints were discovered. Prominent among these equipment was Livescan. Livescan technology makes fingerprinting easier: it is inkless and works quite faster, if compared with other automated devices. The process of using Livescan is as well simple: once a trained technician scans someone’s fingerprints using an Identix TouchPrint 600™ Live-Scan, for example. Then the fingerprints could be electronically transferred to the Department of Justice for record keeping or checking someone’s criminal record. In a similar pattern, facial recognition entails some processes. “Computer facial recognition is a complex and fascinating software accomplishment. “The question, Can a machine recognize faces at least as well as humans?’ is a challenge problem that dates back to early machine vision research in the 1970s, and the problem continues to intrigue and attract researchers from across many disciplines,” (Ware 101). Human beings have different facial looks, even though they seem to resemble one another at times. So, to identify one face from the other has been a critical task for the law enforcement. But in spite of much research in this area, it is still difficult to confidently confirm a technique as the best system to identify individual faces. However, the advent of facial recognition technology helpfully resolves the issue of how could a person’s face being recognized to authenticate his/her identity. Facial recognition technology has passed through two phases, like every other technology. The first phase is termed neural model or computing, which processes images as the way images are processed biologically; for example, in human brains. Neural model can recognize a person’s face through pattern recognition (Chen 200). And its elastic networks can compare facial landmarks that are not practically identical, as a human brain would do. The second phase brings forth a collection of algorithms that can represent and compare faces on the basis of a palette of facial distraction. “The development of this technique is associated with Matthew Turk and Alex Pentland, who demonstrated a process for how facial abstraction images, or characteristic eigenfaces, are generated from a collection of images, and then the faces are expressed as a weighted sum of these archetypal faces,” (Ware 105). Therefore, many face recognition software utilized either of the above-mentioned two models to recognize and compare human faces. The idea of fusion of both fingerprinting and facial recognition has been proved to be the most efficient method of personal identification. This is because it combines the two techniques of identification highlighted above with hope of perfectly identifying who someone is, and differentiating the person from anyone who might look exactly like him/her (Shoniregun 95). The fusion of face and fingerprint by SVM (Support Vector Machine) fusion system improved the overall identification and authentication of human identity. Integrator Fingerprint Face Verifier Verifier Fingerprint image Face Image Discussion The fusion of fingerprinting and face recognition has been adjudged as a more pragmatic approach. The reason for this is that it increases the performance and robustness of identity identification, and it is said to be more practicable than when doing face-only or fingerprint-only identification. A common application of the fusion of fingerprinting and face recognition is the biometric passport. Many countries are issuing biometric passports to their citizens to beat the occurrence of passport fraud (Petersen 45). The biometric passport has both the facial and fingerprint parts. When people cross borders, the authorities there could ascertain the true identity of the passport holders, and thereby control inter-border terrorism, smuggling and crime (Tuyls 10) The fusion of fingerprinting and facial recognition combines the advantages of both identification techniques (Nanavati 30). The fusion strategy has better robustness because it is accurate: there is almost no possibility that the fingerprints or face of a person whose identity being authenticated could be matched with a wrong one (Hammond 115). Therefore, this confirms the enviable reliability of the fusion system. And when needed, the facial and fingerprint images of a person could be selectively retrieved and analyzed. However, there are challenges the users of the fusion strategy would face as per the continuous verification of the results produced by this system. It is impossible to assume the system to be a sum or product of the constituent result of the two combined techniques. For the outcomes of the identification to be reliable, a fusion strategy must satisfy the following conditions: one, the different reliabilities of models joined together in a fusion must be accounted for: this implies that any fusion modality must be a factor of the individual reliability of each of the modality fused together. Two, Any fusion method should be able to handle any lack of observations in one or more of the modalities joined together. This could be as a result of the user’s careless attitude. Three, the older observations should be discarded so as not to influence the reliability of future observations (Zhang & Jain 563). Conclusion For the fact that there are still a few universal applications of the fusion of fingerprinting and face recognition, it is difficult to predict its rate of success. However, theoretically, it works better than each of the identification techniques separately (Zeng 8). And as threats of security expand across the globe, there are possibilities that the fusion strategy would be used commercially; since it is reliable, selective and accurate. References 1. Woodward, John D. Biometrics. New York: McGraw-Hill/Osborne, 2003. 2. Ware, Karl. Biometrics and Strong Authentication. Emeryville, CA: McGraw-Hill/Osborne, 1999. 3. Zhang, David and Jain, Anil. Advances in Biometrics: International Conference. Boston: Birkhäuser, 2006. 4. Tuyls, Pim. Security with noisy data: on private biometrics, secure key storage and anti-counterfeiting. New York: Springer, 2007. 5. Zeng, Wenjun. Multimedia security technologies for digital rights management. Seattle, WA: Academy Press, 2006. 6. Nanavati, Samir. Biometrics: identity verification in a networked world. London: John Wiley and Sons, 2002. 7. Petersen, Julie K. Understanding surveillance technologies: spy devices, their origins & applications. Boca Raton, Fl: CRC Press, 2001. 8. Shoniregun, Charles A. Securing biometrics applications. New York: Springer, 2008.   9. Chen, Ke. Trends in neural computation. New York: Springer, 2007.      10. Hammond, Riad I. Face biometrics for personal identification : multi-sensory multi-modal systems. New York: Springer, 2007.     Read More