Robotics Reference
In-Depth Information
most successful face recognition system at the time of writing can match
faceprints at a rate of 60 million per minute. As comparisons are made,
the system assigns a value to the match on a scale of one to ten. If a score
is above a predetermined threshold, the system's operator is shown the
two images in order to confirm that the recognition system has found an
accurate match.
The primary users of face recognition systems are security and law
enforcement agencies, because faces that are “captured” in crowd scenes
and elsewhere can be compared to those in a database of criminal mug
shots. One of the most important security applications, particularly in
the aftermath of 11 September 2001, is the use of face recognition tech-
nology as a major weapon in the war against terrorism. In 2001 Keflavik
Airport in Iceland became the first airport in the world to announce its
use of the technology to screen passengers.
An innovative use of this technology has been the Mexican govern-
ment's idea of weeding out duplicate voter registrations. It was a popu-
lar practice for voters in Mexico to register several times under different
names in order to be able to cast several votes at elections. Using face
recognition technology, Mexican officials can search through facial im-
ages in their database of registered voters in order to identify duplicates.
The technology was used in the 2000 Mexican presidential election with
considerable success, and was later introduced into local elections.
Another potential application for face recognition technology lies in
ATM and cheque cashing security. Software is able to quickly verify a
customer's face, helping to protect customers against identity theft and
fraudulent transactions.
Veggie Vi s ion
Jon Connell and his team at IBM's Thomas J. Watson Research Cen-
ter have developed Veggie Vision, a computer system that can recognize
each of 150 or so different fruits and vegetables. By using a Veggie Vi-
sion scanner a checkout assistant at a supermarket can speed up their
customers' waiting times and reduce the errors caused when a human
mistakes one type of produce for another. In fact, Veggie Vision is even
better than that. It can be used without a human assistant.
The Veggie Vision systemmakes good use of colour information. But
although there are many obvious differences in colour between different
items of produce, there are also many colour similarities, for example
