Home           Company          Technologies        Products         Solutions         Knowledge         Business         Download         Contact
Search:
 
　

Face Recognition

Near-infrared face recognition system
Face Surveillance and Recognition System
Large Database Face Search System
Face Identification and Verification System
Intelligenet Video
Single camera system
dsfs Multi-Camera Fusion
Video Summarization
Video Retrieval

 

Authenmetric--Quick identity recognition system in large database
     

Technical Background:

  

With today's severe international security measures, a rapid and accurate identification system that can detect identities within a large crowd of people is an important means to protect public security, for national security, judicial, e-commerce and other security systems in related fields.

Face recognition is the most natural, visual, promising way to recognize the biological identity, in conformity to the human physiological visual habits as mankind identify others by their face. With the rapid development of face recognition technology, it is possible to use this technology to quickly and accurately carry out identity recognition from a large database of images.

Face Recognition is sensitive to light, gesture, facial expression and age changes. Facial accessories (such as glasses, etc.), makeup, camera equipment, image quality, and other factors have a great impact on the performance of face recognition, and adds to the difficulty to recognize an image of a person when subjected to notable changes. The diversity in larger databases of facial image sources makes face recognition more difficult to compute accurately and effectively.

 

Systematical outline:

  

AuthenMetric’s quick identity recognition system is able to match people one by one with their photos within a large database, and make a list by order of similarity to recognize the person in photo quickly and accurately. The system supports tens of millions of database inquiries, making full use of existing second-generation ID card photo resources from the public security departments in order to provide assistance to the public security departments.

Notable features:

  
  • Various ways to import images: Face Image database importing, video and photo collection are all acceptable methods
  • Efficient Anti-attack face recognition algorithm: correctly recognizing aging and make-up.
  • Multi-user parallel retrieval: supports parallel recognition processing; providing identity retrieval to many users simultaneously.
  • Massive database applications: : distributed multi-processor system? allows the expansion of over 10 million facial images
  • Swiftness and efficiency: inquiry by single computer of 1 million images databases within two seconds response time, and by Client/Server structure of a computer network group within no more than 8 seconds each time.
  • Flexible client mode: sentinel-style and portable client side mode.
Product description:  
 

The system consists of three parts: client side, server-side and the main servers’ database (Distributed database)

  • client terminal
    A Client terminal program provides a user-friendly interface, responsible for the display of human face figure and output of the matching results with the database. The face image aquisition can be accomplished in two ways: wired or wireless.
  • secondary server terminal
    The running programs on the server mainly handle large data management and enquiry processing, provides data computation services for the client program to compare and match the real-life images ?with the pictures in the database.
  • main server terminal
    Manages the distribution of data to the various subordinate servers to enhance system performance.
System hardware
  1. the host computer, or client side
  2. the main unit group, or affiliated server
  3. the main unit, or main server
system software
  1. client side software
  2. affiliated server software
main server side software

 
Performance indicators:  

Classification

Technological parameters

Match accuracy

90%（image of equal quality）

Match rate

1 million times/second

Match mode

1:N

Face image pattern

support all Windows image files

Face image size

distance between one eye and the other

Face image quality

Clear front image with distance between one eye and the other r＞ 50

operating system

Windows2000/WindowsXP

database

SQL Server/ mySQL/ file system

settings of main unit affiliated server

cpu p4 3.2G，memory 3GB，harddisc100GB；

settings of main unit in client side

cpu p4 3.2G，memory 1GB，harddisc10GB

Settings of main unit in server side

cpu p4 3.2G，memory1GB，harddisc100GB

the number people being checked per affiliated server

1 million person

language

Chinese /English

 
System structure:  
Typical application:  
  • Criminal Investigation
  • comparison with criminal records
  • general population management
  • dedicated population or group management
  • quick identity recognition of monitored people
Face Recognition seamlessly integrates with second-generation ID cards recognition to greatly expand the spreading and application of the face recognition in public security, tourism, aviation, transportation and many other sectors.

 

 

 

 
 
 
 
隐私政策  |  使用授权协议  |  联系我们         
北京数字奥森科技有限公司 @2004-2008 .