Facial Recognition Technique
The brand-new self-organizing computerized technique, invented and developed by HNNY company, makes it possible to recognize people basing upon their photographic images - provided the relevant picture exists in a portrait database.
It takes the system less than 1 second to identify the person, this tiny time span is totally independent of the database volume. Tests have proved that even while dealing with huge databases (containing over 1 million portraits), speed and accuracy of search do not decrease.
How it works: a freshly-taken picture of a person in query is converted to a set of digits that serves the key to search. All the photos in the database have been digitized beforehand; their "old" numbers are compared with the "new" ones, paving way to the identification of the doubtful face.
The system is a self-organizing one. When added to the database, new faces slightly retune the search engine, thus increasing the search accuracy.
As an additional advantage, the possibility of keeping the portrait database in a single data center should be mentioned. In order to have a photo recognized, you need not send the picture itself: its transformation into digits may take place inside the photographing device (camera, cellular phone, etc.). For such purpose, special software should be installed.
After its conversion into digits, the freshly-taken image is forwarded to the database (by means of the Web, phone, etc.), and the search result can be displayed on the screen and/or mailed to the chosen address.
In the most effective way the system may be adopted by companies and institutions which encounter a constant need for rapid facial recognition among large crowds. It seems to be extremely useful in airports, banks, as well as military and law-enforcement facilities. The technique will be particularly valuable for security services.
As soon as somebody approaches a surveillance camera equipped with our system software, the device instantly takes and digitizes his/her photo and sends the relevant range of numbers to the database. In a fraction of a second, the headquarters receive the conclusion: whether the person is an employee (whose portrait appears in the company records) or an unauthorized one.
Using the new system will significantly improve the safety standards and reduce costs of guarding important objects.