Q: What is video analytics?
A: A recent statistic from the Metropolitan Police Service stated that only three per cent of London crimes were solved using CCTV. However, with more than 10,500 public CCTV cameras in London’s 32 boroughs alone, there is huge potential to drastically increase the number of crimes solved through the use of video analytics – not to mention numerous possibilities in the private sector.
So what is video analytics? It is the science of using technology to analyse and manage video, often in large volumes and in multiple formats, from CCTV to mobile phone footage. This transforms video information into intelligence and potential evidence upon which to make decisions.
Video analytics can be real-time: using technology configured to track and provide alerts to specific incidents as they happen; or post event: retrospectively searching for incidents that have already occurred.
Q: What different types of video analytic technologies are available?
A: There are two types of video analytics technology: incident detection and motion detection. Incident detection uses mathematical algorithms to analyse video. It uses tangible parameters such as size and shape of an object as well as the direction of movement to track incidents. Many police services use this technology, including West Midlands and Humberside. This is because the technology is more targeted and provides more accurate results, saving time and money.
Using artificial intelligence (AI) technology, motion detection is less intuitive and you need more skilled operators to ‘teach the system’ in contrast to incident detection, which is more structured and easier to use. Also, AI-based systems can be very sensitive to any movement, often triggering false alerts such as animals walking past which, more often than not, are irrelevant and need filtering out.
Q: What types of application are there?
A: With increasing volumes of video surveillance, there are numerous applications in both the public and private sectors, from general security, to government, retail, transport and financial services. These applications protect people and assets against harm and damage, ideally before events occur. Systems enabled with video analytics can be configured to:
– capture only people lingering longer than a specific time, such as five seconds at a known crime spot while ignoring passers by;
– trigger an alarm if people approach a perimeter fence, while ignoring animals or unusual light anomalies;
– determine the location, speed and direction of travel;
– identify vehicle license plates;
– evaluate how long a package has been left in an area.
Former Chief Inspector of Constabularies and Chief Constable of West Midlands Police, Lord Dear says: “With an intelligent approach to surveillance, I foresee a significant market for private security companies, transport, defence, and indeed any businesses wanting to protect high-value assets. But also I see massive potential for CCTV, liberating the operator so that not every screen needs to be watched.”
Q: What are the benefits of video analytics?
A: With ever increasing amounts of video evidence available, the problem of extracting information has been likened to finding a needle in a haystack. If all UK surveillance footage was analysed manually, the estimated cost would be nearly GB pound 1.25 billion per annum. Video analytics automates the process, saving more than 90 per cent of the cost.
For example, video analytics reduces the time taken to analyse 24 hours of footage down to less than one hour. Also, recent developments in video analytics give the ability to increase the speed and efficacy of searching for specific events or incidents and then presenting the evidence in court to help secure convictions. The technology is saving police forces and private commercial security companies millions of pounds while removing the need for blanket surveillance of law-abiding people.
Q: What does the future hold for video analytics?
A: To date, the focus has been on obtaining video information to tackle crime, with insufficient emphasis on how that information can be used intelligently. There is a significant opportunity ahead for video analytics technologies that are platform-independent to integrate with different IT systems. Such systems can search more rapidly and create more intelligent information on which to base decisions.
Also, video analytics can now transmit information via different means (3G to wireless and satellite) to detect crimes and provide evidence almost instantly. For example, body worn video devices worn by police officers can relay evidence from a crime scene back to a police station and increase the likelihood of securing convictions. This is in contrast to previous scenarios where criminals have not been prosecuted as a result of a lack of evidence.
