Forum for Science, Industry and Business
Sponsored by:     Siemens  n-tv 
Search our Site:

Topic (optional):

 

Home Reports Information Technology Content

Smart turnstile

next article
05.05.2006

 


A biometric security system which had been developed by specialists of the Scientific Production Association “INFORMATION” was awarded the best innovation product title at the 12th International Exhibition "Guarding, Security and Fire Protection MIPS-2006". The system’s peculiarity lies in utilization of a 3D image of the face for identification of a person. Only within a second and a half, the turnstile equipped by this system would create and analyze such image, and take a decision based on received data – to allow or to prohibit the entrance to the person.


It should be noted that although the device does not look fantastic in outward appearance – it is something like an ordinary “three-fingered” turnstile in the underground or a railway station equipped with a small screen on the board, it is arranged interestingly indeed as it identifies persons not by common, so to say visible photographs or three-dimensional portraits but by a special portrait. As a result, the turnstile can hardly be deceived either on purpose or by accident – neither ordinary masks or guises put on by time would manage to cheat it. The portrait for identification is a special one.

The portrait is made as follows: infrared rays are aimed at the person’s face as if the rays were passed through a multitude of parallel slits (like through a fence or railings). These stripes of shadow and light bend and change on the surface of the face, this being done in a strictly individual way. The camera-recorder (also in an infrared band) captures this striped image and transmits to a PC. A PC, in it’s turn, builds a three-dimentional image by line bends with submillimetric precision, identifies it - compares with the database and opens or closes the turnstile. At the same time, it performs additional tasks – for example, it records the employee’s arrival/departure time.

To avoid the system being misled by the changes, which, alas, happen to everybody – people grow stout or thin, and, eventually, grow old – the authors have suggested to use that special control points on the face, which almost do not change with age. These points are the bridge of nose, corners of the eyes and many others - about 60 altogether. By the way, the control points were selected strictly in accordance with science – or more precisely, with the help of anthropologists. Acting in compliance with the method of Gerasimov, a well-known Russian researchers who suggested the way to reconstruct the outline of faces by the shape of skull, Gerasimov’s followers helped to choose the set of points necessary and sufficient for exact identification of the face.

Of course, the system does not perform absolutely reliably – the authors are aware of that. In one case out of ten thousand, the system can make a mistake and let pass an unauthorized person – if his(her) face provides for exactly the same 3D-portrait as one of the employees does, but this is extremely unlikely. As for the authorized employees, the system can make one of a hundred persons pass the test for the second time – but this will only take a couple of seconds. In fact, if a person grows a beard that covers half the face and puts on glasses and pulls the hat over the eyes, the system would definitely give up – and declare the person a stranger. The person would have to unmask.

In any case, the system’s reliability is much higher than that of ordinary turnstiles with magnetic cards or even that of an admission system with a security guard at the entrance. Anyway, such “permit” can not be lost, left at home or handed over to a colleague so that he/she registered him/herself on your behalf. Do not lose the face – and the smart turnstile would appreciate that.

Sergey Komarov | Source: alphagalileo
Further information: www.informnauka.ru

next article

More articles from Information Technology:

nachricht Siemens develops a video solution to uncover leaks in an industrial environment
20.11.2009 | Siemens AG

nachricht ‘Fingerprinting’ RFID Tags: Researchers Develop Anti-Counterfeiting Technology
20.11.2009 | University of Arkansas, Fayetteville

All articles from Information Technology >>>

B2B Search

Product / Service
Company / Organisation

Latest News

Scientists Unravel Evolution of Highly Toxic Box Jellyfish

20.11.2009 | Life Sciences

When good companies do bad things: Examining illegal corporate behavior

20.11.2009 | Business and Finance

UCR plant scientist's research spawns new discoveries showing how crops survive drought

20.11.2009 | Agricultural and Forestry Science

VideoLinks

Event News

Multidisciplinary meeting on Urological Cancers aims to benefit cancer patients

20.11.2009 | Event News

'Golden Age' for clinical psychology in Northern Ireland

20.11.2009 | Event News

New Perspectives in Marine Anti-Fouling Research

11.11.2009 | Event News