Video analytics within CCTV: how intelligent systems maximise security
Video surveillance systems capture a large amount of video, particularly if there are multiple cameras continuously recording 24/7. Most of the footage captured, however, does not get reviewed due to lack of time or personnel. If the site is not monitored, suspicious behaviour may not be detected in time to prevent incidents and security breaches may be missed, leaving the premises vulnerable.
Video analytics, however, can help to monitor footage in real time.
How do video analytics work?
Video analytics, also known as intelligent video surveillance, uses software applications that analyse video feeds, breaking images down pixel by pixel, to identify what is actually happening in the video, and differentiate between people, vehicles or animals in the scene, as well as their appearance and movements. This information can be used as the basis upon which to perform certain actions, e.g. to decide if keyholders should be notified, or if an audio warning should be issued.
The algorithms within the software can be used to:
· Detect a person or vehicle entering a restricted area, using line-crossing technology
· Identity traffic movement (such as traffic travelling in the wrong direction)
· Detect a person or object loitering in an area
· Object tracking
· Count and/or classify objects
· Facial Recognition
What are the benefits of video analytics?
· Triggering real-time alerts
Video analytics can trigger an alarm to a monitoring station or keyholder as soon as unusual behaviour is observed. This could include an alert for a person or vehicle crossing a pre-defined unauthorised area, a count-based alert when a certain number of objects are observed within a given period or even facial recognition alerts (using deep learning technology). This real-time monitoring can help to prevent crime, rather than reacting after an incident has occurred.
· Reduce false alarms
If the site is remotely monitored by an Alarm Receiving Centre (ARC), alarms can be triggered for movement by PIR (passive infra-red) detectors. Traditionally, any movement would alert to the ARC and have to be reviewed by an operator. Alternatively, many modern cameras have ‘false alarm filtering’ technology, which uses video analytics for recognition and identification of objects in the frame. For example, distinguishing between wildlife and human activity. The monitoring station would then be able to establish if movement was human, rather than an animal or greenery rustling in the wind, without having to manually review the footage. This can help monitoring centres establish whether the video captured is suspicious quickly, and take the correct follow-up action.
· Business Intelligence
Not only can video analytics be used for crime detection and prevention, but can also be used for business intelligence purposes, particularly useful in retail settings. For example, people counting solutions can indicate the busiest periods for a store, which in turn could streamline staffing levels. Heatmapping technology can indicate where people spend most time in the store, which could help with merchandising.
Watch Systems Ltd can help design the most suitable video surveillance system for you and your premises, utilising cameras with intelligent video analytic solutions for enhanced security. To learn more about video surveillance and video analytics, get in touch with our team here.
