Video Analytics: Helping businesses to re-open in the wake of COVID-19

National Governments across the globe are absolutely determined to contain or otherwise delay the spread of the Coronavirus which, at the time of writing, has seen 1.5 million individuals worldwide infected and resulted in 90,000 deaths. Those Governments are resolutely locking down whole cities. Those places where people congregate in large numbers for leisure activities, such as retail outlets, restaurants, pubs and clubs, are closed until further notice, with just essential shops (ie supermarkets) and chemists permitted to remain open.

At some point - hopefully not too far into the future - the number of people testing positive for COVID-19 infection will begin to reduce. At that juncture, thoughts will turn towards the business community (among others) and how companies might safely re-establish their commercial activities. Inevitably, social distancing rules will have to remain in place with a phased return to normality. This is likely to mean that only a specified number of people will be allowed to congregate in a controlled area at any one time.

There are so many environments where complying with this limitation will be a challenge. The aforementioned pubs and clubs, for example, will be desperate to re-open at the earliest possible opportunity, but in order to do so, management teams might have to employ additional door staff just to count the number of people who enter and leave the premises. For companies already struggling to survive at present, the cost of hiring these extra members of staff could be prohibitive.

People-counting video analytics

Fortunately, there's a simple solution readily available which can be quickly deployed and will prove to be far more cost-effective and accurate than employing someone to do the same job. It comes in the form of video surveillance cameras supplied pre-loaded with people-counting video analytics

In practice, such analytics are able to provide real-time information on the net number of people on the premises at any particular time. The software is able to simultaneously count those individuals entering and leaving a given premises. The resultant data can be displayed in an easy-to-understand format on smart phones, laptops and PCs, in turn providing managers with an early warning whenever permitted numbers are close to the allowable limit.

System integrators will no doubt be able to identify many other social environments where people counting video analytics will help to ensure that social distancing rules are implemented and verified. We don't yet know how these rules will evolve when the immediate crisis is over, of course, but Shopping Centres, art galleries and museums, etc are just a few more examples of where the solution can be effectively deployed.

Although developed to help retailers gain a greater understanding of customer behaviour, video analytics capable of identifying occupancy levels provide a powerful tool for identifying whether or not too many people are congregating in specific areas.

As an example, Artificial Intelligence (AI)-Crowd video analytics are able to provide an estimation of the number of people present in crowded areas where they might be expected to move slowly or stop, while AI-Overcrowd will generate an alert if the occupancy of an area exceeds a specified threshold. As is true of people-counting analytics, this solution can be quickly deployed and used ‘out-of-the-box’ as the software can also run on-board those cameras complete with open platform chipsets and sufficient processing power.

Many surveillance cameras will support two-way audio and, as such, enable operators to communicate with visitors that they need to move out of a given area.

Intelligent analytics

In addition to these specialist video analytics applications, a large number of IP network cameras now include as standard various forms of license-free analytics. Some are equipped with a suite of Intelligent Video Analytics, which includes enter/exit direction, tripwire and face detection. Combinations of these various forms of video analytics can be configured to form a cost-effective 'digital barrier'.

With the threat of increased criminal activity, Control Room operators will need all the help they can muster to visually verify any suspicious activity and, if necessary, direct security personnel to deal with any incident. The earlier they can do so the better. A new range of cameras will soon be launched to market equipped with deep learning-based video analytics. They will ensure that operators are not frustrated by ‘false positive’ alerts which, depending on the environment, might be generated by existing forms of video analytics, as well as the false alarms often generated by intruder and perimeter protection sensors.

These cameras will be able to simultaneously detect and classify various object types, including people, vehicles, faces and license plates. Most importantly, the software involved ignores video noise, waving trees, moving clouds and animals, all of which might normally be the cause of false alarms.

Importantly, the AI algorithms built into these new cameras will also provide greater people-counting accuracy. 

Innovative solutions for 2020 and beyond

Not too far into the future, we will see the introduction of AI applications designed to enhance video surveillance systems to a level which, at this point in time, might not seem possible. Many of the innovations on the horizon will potentially give greater assistance to businesses as they seek to operate in a world where social distancing will remain a very important issue for some time to come.

These advances in video surveillance technology will also provide security personnel with powerful tools to keep one step ahead of criminals. Indeed, the criminal fraternity are unlikely to restrain themselves at this challenging time. They will most likely seek to take advantage of a situation where many business premises will be unoccupied and, therefore, offer rich pickings to looters and thieves. There's certainly no shortage of high performance cameras available which will allow Control Room operators to detect any suspicious activity.

For some time now, video surveillance system manufacturers have been striving to help end users achieve added value from their solutions. COVID-19 has made it even more important that this trend is maintained. Fortunately, our industry has no shortage of ingenious design and development engineers who will help us rise to the challenge and absolutely ensure that we contribute positively towards companies once again being ‘open for business’ as quickly as possible.

Uri Guterman is Head of Product and Marketing for Hanwha Techwin Europe