Scroll Top

Edge AI

Face detection & recognition

edge ai

Face detection and recognition system

Robust solution for people identification, queue management & improving security in commercial facilities and public transportation.

How does it work?

Want to see it in action? Check out the movie and contact our sales to request PoC.

The idea behind our solution

Face detection and recognition algorithms and presence control systems are widely known to entrepreneurs, open source developers and end users familiar with mobile Face ID technologies. However, there are still business cases that are not covered by commonly available solutions.

What challenges do you want to overcome?

  • Do you want to replace RFID cards or fingerprint readers with contactless technology?
  • Are you looking for a reliable system for automatic and contactless authentication?
  • Do you want to increase security for passengers or employees at a managed facility?
  • Do you want to reduce the latency associated with cloud computing or transmission of large files?
  • Or do you have other challenging issues that can be solved with Edge Computing and AI?

A solution powered by Edge AI

Edge AI

Edge AI computing

The facial detection and recognition system runs on our custom Raco Edge AI Gateway. All operations are performed on the device without the need for cloud computing. Raco has a built-in camera and can be extended with any other model connected via CSI or USB ports.

Key benefits

szybka obsługa

Fast service

Our solution is fast and reliable - the authentication and detection process takes milliseconds.

zgodność z rodo

RODO compliance

There is no need to store sensitive data - the system can run on the device, and identification is fully anonymous.

inteligentny system

Smart system

The system adapts to changes in appearance, recognizing a person wearing glasses, a hat or a new hairstyle.

łatwa integracja

Easy integration

Integration with external systems is possible via API and network connection via Ethernet, Wi-Fi or LTE.

wydajność

Time efficiency

Adding new people to the registry is quick and only requires taking a few photos to train the artificial intelligence model.

Business scenarios

rozpoznawanie twarzy

Employee facial recognition

Identifying factory workers during shift changes and recognizing corporate workers during peak hours to prevent long queues.

Zapobieganie kradzieży tożsamości w transporcie

Preventing identity theft in transportation

The system prevents security breaches and fraud by detecting people using a stolen ID card or an unauthorized ticket discount.

Zarządzanie czasem i kolejkami

Time and queue management

Generate savings for the company by better managing the time of employees waiting in line for authentication to start their shifts.

Zwiększenie bezpieczeństwa personelu

Increase staff safety

Replace manual identity verification with a reliable solution that automatically recognizes authorized individuals.

The process of creating a custom AI model

product creating process

How do we operate?

potrzeby

Getting to know your needs

As a Raspberry Pi Approved Design Partner, we have the necessary expertise to provide advice on IoT technology and deliver a solution tailored to your specific needs.

Model-ai

Training the AI model on your data

The artificial intelligence model must be trained using customer data, according to the principle - the better the training data, the better the model's performance. The effectiveness of the final solution depends on the environment, lighting and the way data is acquired for AI training.

Wybór sprzętu

Equipment selection and device design

Each project is priced individually after the workshop. Development of a ready-to-use solution requires preparation of a custom AI model and selection of appropriate hardware, choosing from Raco Edge AI, NVIDIA Jetson or others.

Technical aspects - types of cameras

access handling

With access support

For face recognition, we can use a standard camera used to monitor objects. However, such a device does not recognize whether the image is three-dimensional or flat, which means that the system will not work in all security situations.

Advantages:

  • Cheaper equipment
  • Faster implementation
unsupervised access

With unattended access

To make the system fully autonomous, we need an infrared-based solution similar to those used to unlock smartphones. The AI algorithm recognizes a real, three-dimensional face and illuminates it with infrared light, which is invisible to the human eye.

Advantages:

  • A safer solution
  • Harder to fool than a device that does not recognize image depth
  • Gatekeeper supervision is not required

Required equipment:

  • IR illuminator – a small lamp emitting infrared light needed to illuminate the entire face

  • IR camera
    , which will verify what the illuminator displayed and transmit this information to the Edge AI model, which will determine whether we are dealing with a real person

Photos taken with the IR camera at all stages of face recognition with a smartphone – the process takes about 1 s

IR camera

From left:

  • standard image seen with an IR camera, image seen without the use of an illuminator,
  • the illuminator illuminates the entire frame (just like a flash when taking a photo),
  • illuminator casts random “dots” on the image,
  • illuminator casts random “spots” on the frame.

Empower your project

Unlock the full potential of your technical aspirations!

Contact our experts to explore the possibilities and drive your projects to a new level.

Empower your project

Discover the full potential of technology with experienced engineers.