Published on March 5, 2014
By Ravindra Das President Apollo Biometrics, Inc.
•Facial Recognition is one of those Biometric Technologies which most people can associate with. •We all have a face, and one of the best examples of facial recognition are the photos in the post office, and law enforcement website. •But unlike the other Biometric Technologies being used today, Facial Recognition is subject to huge public scrutiny
•This is so because Facial Recognition is very prone to privacy rights violations as well civil liberties violations. •Also, Facial Recognition can be used very covertly without the knowledge or the consent of the individuals using it. •Also, Facial Recognition has its fair share of scientific flaws as well
•A Facial Recognition can be spoofed very easily: Suppose it takes a template of a grossly overweight person, and takes the template of the same person after massive weight loss. •The Facial Recognition system will not be able to recognize this individual. •But, it is not just weight loss which can spoof a Facial Recognition System.
•The presence and removal of facial hair, as well as the presence and removal of other objects can also spoof a Facial Recognition System. •Examples of such objects include hats, sunglasses, as well as switching from contact lenses to eyeglasses, and vice versa. •But despite these weaknesses of Facial Recognition, its key advantage is that it can be used in large scale identification applications.
•For example, Facial Recognition is used in the e-Passport infrastructure of many countries, and used for large scale identification applications throughout the world’s major international airports. •In these regards, Facial Recognition can be deployed as a fully automated or semi-automatic system, where human intervention is required (and is the preferred method).
•Facial Recognition technology relies upon the features of the face which are determined by genetics. These include: *The ridges between the eyebrows; *The cheekbones; *The mouth edges; *The distance between the eyes; *The width of the nose; *The contour and the profile of the jawline; *The chin
•It should be noted that Facial Recognition is purely a non- contactless system. •To start the process of raw image collection, the individual must first stand in front of the camera, or under covert conditions, their face will be captured via a CCTV camera with Facial Recognition software embedded into it.
•Once the raw images are collected, the facial data is then either aligned or normalized to further refine the raw images at a much more granular level. •The refinement techniques include: *Adjusting the size of the face to be in the middle of the pictures which have been taken; *Adjusting the size and angle of the face so that best unique features can be extracted later and subsequently converted over into the verification and enrollment templates
•The entire process of facial recognition starts with the location of the actual image of a face within a set frame. •The presence of the actual face can be sensed or detected from the various cues or triggers, such as: *Skin color; *Any type or kind of head rotation; *The presence of the facial or head shape; *The detection and presence of both sets of eyes in the face.
•Some of the challenges of Facial Recognition include: *The identification and the differentiation between the tonality of the skin color and the background; *The various shapes of the face, multiple faces of the same individual may be combined to create an overall, composite face.
•All Biometric devices contain of a capture device to get a picture of the raw, physiological image, and to extract the unique features for template creation. •But the real technology behind any Biometric Technology are the mathematical algorithms which are employed.
•With Facial Recognition, there are three primary types of mathematical which can be used, depending upon the vendor. •These algorithms are: *Principal Component Analysis; *Linear Discriminant Analysis; *Elastic Bunch Graph Mapping
•Principal Component Analysis (PCA) dates back all the way to 1988. •In this mathematical algorithm, a concept known as “Eigenfaces” is utilized. •Eigenfaces are just 2-Dimensional spectral facial images, which are composed of grayscale features
•Literally hundreds of Eigenfaces can be stored in the database of a Facial Recognition system. •When the facial images are actually captured, this library of Eigenfaces are placed over the facial images, or superimposed on top of one another. •The variances found in the superimposition between the facial images and the Eigenfaces are calculated, and different weights are assigned.
•With PCA, the end resultant is a 1-Dimensional image of the face. •In terms of mathematics, PCA is merely a linear transformation in which the facial images get converted over into geometrical coordinate system. •The biggest disadvantage of PCA is that a full image of the face is required. Thus any changes in the facial features requires a full recalculation of the Eigenfaces.
•Linear Discriminant Analysis (LDA) projects a face into a vector space, with the idea behind this is to cut down on the total number of facial features which need to be matched. •The basic mathematical theory behind LDA is to calculate a single raw data point from a single raw data record, and from this linear relationships are then calculated.
•From the linear relationships derived, pixel values are derived, and plotted. •The end resultant is a computed face image, which is also called a “Fisher Face”. •One of the biggest advantages of LDA is that it take into account the various lighting differences, and the various types of facial expressions which can occur.
•Elastic Bunch Graph Mapping (EBGM) looks into the non mathematical relationships of the face. •Some of these relationships include lighting differences, and the differences in facial poses and expressions. •With EBGM, a facial map is created, with various nodes at the landmark features of the face which are the eyes, lip edges, and eye edges.
•Facial Recognition applications have grown extensively since the days of 9/11 •Right after 9/11, the financial value of many Facial Recognition vendors grew, because Biometrics was all the hype •After the hype was over, and once the public realized the shortcomings of Facial Recognition, stock prices of these vendors nosedived
•But in the last thirteen years, Facial Recognition has come long way in terms of market dominance. •The biggest market application for Facial Recognition is that of the international airports. •For example, Facial Recognition cameras can be placed literally at all of the strategic points of entry and exit, and within minutes, potential suspects can be identified and questioned
•Also, Facial Recognition is being used quite heavily in the e- Passport infrastructure. •The e-Passport is the “electronic version” of the traditional paper passport. •It consists of a smart card which can contain the templates of various Biometric modalities.
•These templates include those of Fingerprint Recognition, Iris Recognition, and Facial Recognition. •The e-Passport infrastructure has been adopted by over 135+ countries, worldwide.
•Another major application for Facial Recognition is that of covert applications, especially in that of large public venues and military theatre of operations. •Facial Recognition technology can is now being implemented into CCTV technology, thus making it a very powerful multimodal security solution.
•Another major application for Facial Recognition is that of covert applications, especially in that of large public venues and military theatre of operations. •Facial Recognition technology can is now being implemented into CCTV technology, thus making it a very powerful multimodal security solution. •Of course, Facial Recognition has even garnered applications and controversy, with the likes of Facebook and Google.
•Probably one of the best known covert applications of Facial Recognition occurred in January, 2001. •At Superbowl XXXV, the Tampa Bay police used Facial Recognition to identify criminal suspects-subsequently, 19 people were apprehended. •Facial Recognition has also started to make its mark in the wireless world, where it is being used on Androids and iPhones to unlock specific mobile software applications.
•Facial Recognition is a non-contact technology. This means that the end user does not have direct, physical touch with the Facial Recognition System. •Facial Recognition does not require the direct cooperation of an end user. •Facial Recognition can be used very covertly, thus making the most uncooperative end user into the most cooperative one without their knowledge. •Facial Recognition is very well suited for mass identification applicationssuch as that of an international airport setting.
•Facial Recognition is still not a perfectly developed technology. •It still suffers from many hurdles, and still struggles to perform satisfactorily in certain types of applications. •Facial Recognition suffers greatly from variables in the external environment, such as differences in lighting, and any other objects obstructing the end user’s face, such as a hat, sunglasses, or even long hair. •The biggest disadvantage thus far is that Facial Recognition still suffers greatly from privacy rights issues and claims of civil liberties violations.
•Facial Recognition will always be plagued by privacy rights issues and cries of civil liberties violations. •But over the last decade, Facial Recognition has improved greatly in terms of technological advances. •Facial Recognition will not be used for small scale applications, but rather, it will be used for large scale verification/identification scenarios. •Also, with the current threats of terrorism, Facial Recognition will be used very covertly with CCTV technology in order to apprehend terrorist suspects, and its greatest worth will be proven in the battlefield.
•Contact Information: •E-Mail: firstname.lastname@example.org •Phone: 312-803-0263 •Address: 330 North LaSalle Street, Suite 4925, Chicago, IL 60654
Facial Recognition: Limited Application in ... facial recognition technology has ... prominent facial recognition application on the market, ...
... recognition technology and applications for facial image ... face recognition technologies and applications for enterprise ... Cognitec Systems ...
... on the Animetrics' facial recognition technology ... Technology Now on Commercial Market ... Science magazine mentioned the ...
... recognition technology and applications for facial image ... About Cognitec. ... market-leading face recognition technologies and ...
... Image recognition market categorizes the global market technology, by application, ... from MarketsandMarkets ... Facial Recognition: Market ...
face recognition technology is being used to combat passport ... In face recognition applications, ... Science, Technology, ...
... In this article about biometric systems the general idea is to use facial recognition ... Recognition in this technology ... recognition application ...
Image Analysis for Face Recognition ... biometric technologies, facial recognition and middleware are projected to ... Face recognition market [9 ...
Advancing the Science of Biometrics Identifying individuals through facial recognition ... automated face recognition technology in law ...