We are sensitive to give exact solution to our customers.

Semi-centralized solutions

In this configuration, each branch runs face detection and generates Ayonix person-specific biometric data (AFID – Ayonix Face IDentification). These AFID data is sent to the central server and central server gives the matching results.

In Traditional solutions, whole live camera images are sent to the server.

In Centralized solution, only face area is sent to the server by reducing the number of data to be sent hence a fast matching processing speed.

In Semi-centralized solution proposed here, only afid data is sent. Therefore, network load is even reduced.

Solution Architecture

What is Local, What is Central Server?

Local is the customer store or the place where camera installments are done. Here, cameras are directly connected to computer. Computer runs face recognition.

Central server is customer data center or customer server room (IT Room) where customer has its own company servers, data storage.

A VPN (Virtual Private Network) is used between Local and Central server. Therefore, data transfer is done in a very secured way.

Here is a comparison chart of different solutions

Traditional systems Centralized solutions Semi-centralized solutions
Data size per second 8MB/second 200KB/second 80KB
Maximum cameras 20 cameras/layer 200 cameras/layer No limit
Matching speed/camera 4-5 seconds 250ms 100ms
Maintenance period Once a month Once a year Once a year
Cost Expensive Cheap Cheap
Service expansion Difficult Easy Very easy
Business scale Small scale Mid-scale Large scale


  1. No expensive hardware is needed in customer side
  2. Server loading is balanced between local and central server side
  3. Minimum network configuration is enough for running
  4. Maintenance is easy and simple.

Suggested applications for this architecture

Franchising stores such as Pizza hut, Starbucks, Mc. Donald, Seven Eleven, Cargo companies

Face city surveillance (Especially for law enforcement at airport, city centers)