Ayonix has various face recognition products and services, platform/network independent architectures.

One of these is the embedded camera solution.

Ayonix Public Security (APS) utilizes Windows service and an application layout controls whole service by a user interface. In this structure, we use following configurations

CPU Intel Xeon
OS Windows 2012 Server
Max. cameras Up to 2 cameras(8MP) orUp to 4 cameras(3MP) orUp to 8 cameras(1.3MP)

Number of cameras to be connected to one server machine is directly related to machine specs. The more cameras we use, the more hardware costs increase

Ayonix embedded camera solution finds an answer for this problem. We developed a service which runs directly inside the Axis camera



A comparison of traditional and Ayonix solutions is given below:

Traditional approach Ayonix-Axis Embedded solution
Face recognition server takes all camera stream Axis camera finds face, gives gender, age information, provides biometrics data
All works are loaded to Face server All loads are done in each camera
Server load depends on number of cameras Server load is independent of number of cameras
Adding one camera to the system is difficult Adding n number of cameras is easy

Comparison of server configuration is given as:

Competitors Ayonix
CPU Intel Xeon Intel Xeon
OS Windows 2012 Server Windows 2012 Server
HDD 100GB 100GB
Max. cameras Up to 2 cameras(8MP) or Up to 4 cameras(3MP) orUp to 8 cameras(1.3MP) 10,000
Recognition time 200ms (two cameras)450ms(four cameras)800ms(Eight cameras)No more camera support 100ms for any number of cameras
Matching speed 650ms(2 cams)/1million records1250ms(Eight cameras) 550ms/1million records

Not more than 550ms for more cameras

Network delay YES NO
Real-time match NO YES
Configuration Difficult Easy


Axis P1367