Video Content Analysis (VCA) is changing the face of CCTV security, says Andrew Hunter, Product and Solutions Director at Chubb Fire & Security, pictured.
There is no question that CCTV has revolutionised the world of security. Since it was first used in the 1940s, it has evolved enormously. Now with the emergence of Video Content Analysis, CCTV as a security tool has become powerfully intelligent and is changing the face of CCTV.
Video Content Analysis (VCA) has taken CCTV security to a whole new level. It has given security professionals the ability to assess events, identify vehicles, people, animals, and monitor meaningful events in movement, while cancelling out non-related events such as tree and branch movement.
While ‘motion detection’ isn’t anything new – it’s been around for 30 years or more – Video Content Analysis is much newer. Over the last five to ten years, this technology has been embedded into CCTV cameras or Network Video Recorders (NVR) and enables the system to work automatically to detect and analyse temporal and spatial events. More recently, additional analytics can be applied post-recording in the cloud to allow the application of Artificial Intelligence to constantly evolve, refine, and improve a security system.
Originally, Video Content Analysis was primarily used in high-risk or high-secure environments, but due to the reduction in cost, it has become much more widespread and is being used in all applications, from large scale commercial and industrial scenarios to peoples’ homes.
The video data generated by IP CCTV cameras gives security teams a better situational awareness. They can use the data to process, categorise, and analyse objects and activities. It can identify temporal shapes, spatial events, and even direction of travel. These powerful capabilities are significantly changing the face of CCTV. Here at Chubb, the level of security we can provide is far more powerful, accurate, and efficient than it has ever been, and the pursuit of incremental improvements remorseless.
In a real-life scenario, for example, Video Content Analysis can pinpoint a security breach by analysing a person’s movements. It can tell the difference between a routine ID card swipe at reception and an intruder jumping a barrier or breaking down a door.
The system is also intelligent enough to determine security breaches by facial recognition. For instance, an alarm would trigger if an authorised person went through a door but was immediately followed by someone not identified.
Video Content Analysis enables us to solve our customers’ biggest problems much more efficiently. If an incident occurs, we no longer need to dedicate lots of manpower or time to searching through event footage. The system will find the actual event almost instantly because it ‘flags’ moments by analytical event type. With this automated system, we also eliminate any human error that could potentially occur such as missed details which may impact on incident response time.
The automated system also gives us a much higher degree of accuracy when determining whether an incident is genuine or not. Our central video monitoring station has been able to reduce the number of false alerts by up to 80% by implementing video analytics. Assessing the validity confirmation of incoming alarms and only presenting the operator with genuine people or vehicle activations is much more efficient for the customer and us.
The other benefit to using Video Content Analysis is a reduction in the amount of hard drive storage you need. The system will only record and keep a genuine event, which ensures storage isn’t filled with needless hours of video content. The reduced need for hard drive space can mean there are fewer servers, which leads to lower power consumption, saving energy and money. We are seeing energy consumption and storage space requirements emerging as a key priority for many of our customers when selecting their system.
While CCTV is synonymous with security, Video Content Analysis can be utilised much more widely. For instance, many of our customers use mobile phone notifications to inform them about relevant events which they require immediate attention. This is particularly useful because it allows the customer to continue with the day-to-day workload while receiving a holistic view, in real time, their entire business and any interventions they are required to attend.
For example, a customer of ours, a major livestock farmer, uses the analytics to monitor their cattle when they go into labour. Because labour can occur at any time – day or night – the farmer can configure an alarm alert when motion is detected from a cow. Cows begin to pace around in circles before giving birth, so when this starts to happen, the farmer can be with his cattle to care for them at the critical time and not have to stay up all night in anticipation.
The future
More and more solutions are premised on multiple analytics providing a series of events which would cause an alarm. Last year, Chubb provided a major petroleum retailer with a platform that recognises if people are on the forecourt and whether they are safely dispensing fuel. This was achieved by multiple analytics being applied simultaneously from smoke detection through to movement and pump impact alerts.
As I mentioned at the outset, CCTV technology is constantly evolving. Over the next couple of years, I expect to see further advancements in self-learning video analytics. This will help to increase the accuracy of false alarm detections and only present genuine alarms, allowing organisations to react more efficiently to different scenarios.
Fast-changing capabilities will also be a significant feature in the coming years. During the pandemic, Video Content Analysis enabled businesses to count the number of people who entered their buildings and detecting whether they were wearing a mask, all designed to minimise the spread of covid-19.
Should other unknown requirements become necessary in the future, then I’m sure other iterations to these analytics will be available through development.
Visit www.chubb.co.uk.
