Detect. Identify. Verify.

Our advanced facial-recognition search engine enables CloudVis and SmartVis to detect and match subjects quickly against a watchlist of known individuals.

The technology is invaluable for fighting terrorism, organised crime or tackling front-line security issues.

It can quickly and accurately identify individuals in non-consensual situations and uncontrolled environments and has been selected by a major European government as the basis for developing bespoke applications.

How does it work?

Our face-recognition search engine is able to work with images and video streams from a wide range of visual sources – CCTV, body cameras, mobile phones or portable devices.

All sightings are then sent to the face-recognition engine, which classifies them by gender and checks their identity against a watchlist.

If a match is found that exceeds a set security level, the results can be presented to an operator.

The likelihood of it being a true match can be emailed to security staff, complete with links to the relevant section of video for further investigation.


Our face-recognition technology can create watchlist entries from faces found in live or pre-recorded video.

Faces can be analysed from any angle, including profile, to ensure the best chance of obtaining reference data and successful identification.

The watchlist takes a short video clip to increase the amount of data available to the matching search engine, resulting in more reliable identification from multiple angles.


Deep neural networks and unique algorithms give the face-recognition search engine unprecedented speed and accuracy. 


When facial recognition is used in conjunction with our other award-winning analytics technology, it is possible to track people with covered or obscured faces.

So if suspects remain on video up until the moment they remove a helmet or mask, the potential is there for a second chance to make a successful identification.

Operators can also check whether individuals match resembling images by using advanced image-comparison features.


Our face-recognition search engine can be applied to live streams or pre-recorded video.

Large volumes of archive video can be reviewed faster than it would take in real time.

It is possible to extract an image of each unique face from a live stream and check against a watchlist or search an archive of video for specific individuals.


Our new deep learning-based facial-recognition algorithms are a massive improvement on the technology available even a year ago – and they will get even better.

They are vastly superior to a human operator performing unfamiliar face-recognition tasks, such as a security guard on a checkpoint or an operator forensically browsing video for suspects.


Poor illumination, shadows and non-frontal poses drastically change the appearance of faces, making consistent recognition challenging.

Best performance is achieved when individuals are within 30 degrees of front-facing and have a minimum of 50 pixels between the eyes.

Our face recognition will still work outside these conditions, and in poor lighting because the technology is based so closely to infra-red imagery.

When the performance is degraded, though, we recommend that face recognition be used as an aid to a human operator going about their daily tasks.

The data to be reviewed can be vastly reduced and prioritised in these scenarios with likely identities sent to an operator, who makes the final decision.


Our facial-recognition analytics are based on the latest in computer-vision research and are being constantly enhanced with additional data.

The current set of facial-recognition algorithms and gender-classification algorithms are Digital Barriers’ first commercial product based on deep learning. Others will follow.

Performance improvements and upgrades are included with the cloud service, making it very simple to benefit from continual development. 


There are business opportunities to be gained too from our face-log technology, which reliably captures each unique face passing a camera, extracting key demographic data.


Our advanced automation tools dramatically reduce operator interaction; lowering costs and improving accuracy.

As a result our face-recognition technology has already received significant interest from alarm-receiving centres, CNI facilities, commercial sites, police and the military.

Examples of how and where we deploy CloudVis’s face-recognition technology:

  • Secure borders
  • Defence
  • Energy and utilities
  • Facilities and site security
  • Safe cities and public safety

Our range of facial recognition solutions

Digital Barriers facial recognition technology is integrated into our CloudVis and SmartVis solutions.

Our SmartVis Face Mobile solution is a smartphone app that enables users with a handheld IOS/Android device to match subjects against a watchlist of known individuals.

It works quickly and accurately by capturing real-time images and video of people’s faces and then comparing them against selected watchlists.