Biometric face recognition works by using a computer to analyze a subject's facial structure. Face recognition software takes a number of points and measurements, including the distances between key characteristics such as eyes, nose and mouth, angles of key features such as the jaw and forehead, and lengths of various portions of the face. Using all of this information, the program creates a unique template incorporating all of the numerical data. This template may then be compared to enormous databases of facial images to identify the subject.
Good biometric software then produces a number of potential matches, rating each based on a numeric score of how similar the match is. When multiple images are used, the accuracy of biometric readings increases greatly, a fact which has provoked the assembly of massive databases, particularly on key figures such as terrorists.
Biometric face recognition is currently used in a handful of American airports, and was used at the 2001 Super Bowl to guard against the perceived threat of a terrorist attack. Nineteen individuals were flagged at the Tampa Super Bowl as having criminal records, but upon closer examination all had only minor infractions on their record.
A number of groups are looking at using this technology in the near future. Airports in the United States and other nations are moving towards incorporating biometric face recognition systems throughout their workings, as the sheer amount of traffic and high potential for terrorist targeting make them an ideal choice. A number of banks have begun test programs outfitting their autoteller machines with biometric face recognition programs, to offer instant cashing of checks without the need for a human teller.
In Britain and other nations which have a history of video surveillance, the transition to using biometric face recognition is contested very little. In the United States, however, which has a strong historical aversion to technologies seen as undermining privacy, there is a major battle being fought between proponents of biometric face recognition systems and its outspoken opponents. Most people who oppose the integration of these systems into everyday environments do so on the basis of civil liberties. They hold that such pervasive identification of who you are — in essence tracking your movements every time you enter a space controlled by a biometric system — violates fundamental rights to privacy and opens up the potential for serious abuse.
The primary use at the this time for biometrics remains for medium-security access to controlled environments. Acting as a replacement for card-keys or thumb-prints seems to be the most obvious use in the near future. At present, even discounting privacy concerns, the technology does not appear to be accurate enough to ensure its adoption for high-security by the world at large. Without a large, centralized database of terrorist photographs, the major argument in favor of biometric recognition lacks much of its force. Though there is a push to assemble such a database, it will undoubtedly be many years before it is sufficient to make biometric face recognition more than an interesting extra security measure; in the meantime it will likely serve as an addition to, but not replacement of, human involvement.
Face Detection & Recognition
The general statement of Face Recognition problem can be formulated as follows: Given still or video images of a scene, identify or verify one or more persons in the scene using a stored database of faces. The solution to the problem consists of solving several subtasks in a sequential manner: face detection, face normalization, and face recognition.
Most existing face recognition systems have limited the scope of the problem, however, by dealing primarily with frontal views, neutral expressions, and fixed lighting conditions. To generalize face recognition systems, one has to look at recognizing faces in varying environment and focussing on to develop methods which deal with different expressions, illuminations, and pose variations.
Majority of the proposed methods use image gray level values to detect faces in spite of the fact that most images today are color. As a consequence, most of the methods are computationally expensive and some of them can only deal with the frontal faces with little variation in size and orientation. To solve these problems, we have investigated the color-based face detection methods.
Thus, there is a developed a new architecture for face detection in color images. Based on skin locus and successive detectors, the method allows high efficiency under drastically varying illumination conditions.
Related Online Articles:
No comment yet. Be the first to post a comment.