The use of camera surveillance, especially in urban areas, is increasing at a rapid pace; as city population densities grow so does the need for more effective surveillance which in turn means that huge amounts of raw data need to be analysed quickly and effectively. To do this manually would place huge strain on our human resources – hence the development of ever more intelligent video analytics.

Despite technological advancements, such as higher optical zoom capabilities, UHD footage, smart IR, etc., without having a quick and effective system to interpret the data accurately it’s all for nought. This is especially challenging in high traffic areas such as subways, busy intersections, traffic control scenarios, etc. And that’s not even taking into consideration how complicated setting up networks would be (the surveillance camera needs to be connected to the network, have access to power and be able to store this huge amount of data to a central storage device that can handle the capacity needed).

Video analytics make use of algorithms based on preselected settings that filter the data coming in from the cameras to provide you with real-time analytics. You can also select specific automated responses that can be performed without human assistance: for example the latest XProtect licence plate recognition (LPR) software from Milestone allow the users to control access to protected areas and premises through an automated system: only vehicles with licence plates that match the database will be granted access. 

There are a variety of features and applications that video analytics can be used for that are used to enhance the human operators’ vigilance and supervision tasks, for example, people counting, facial recognition and detection, line crossing, intruder detection, etc. But the main two benefits of video analytics is that you can set alarm triggers and they help you search through the mountain of stored footage for specific events/footage files.

Alarm triggers: When analytics are applied to real-time video footage you can apply behavioural rules that when violated trigger alarms. The most common of these are people counting, face detection, intruder detection, etc. Milestone and NUUO provide various analytical option and alarm triggers such as point of sales (POS) solutions (Read More). Manufacturers, such as VIVOTEK and Uniview, are also producing cameras with edge analytics built into the IP cameras themselves.

Assisted video search: The NVR will record events where the alarms have been triggered by providing video clips or snapshots of the event that help catalogue the events more efficiently so that when you need to recall old footage it’s easier to find.

Future: Developers are constantly researching ways to improve surveillance technologies. And although video analytics is an amazing tool for monitoring and controlling things in the present as well as acting as a smart archive system of past events, could it one day perhaps be used to predict events based on patterns and behaviours recorded over certain periods? Only the future will tell…

Source: editorial was based on an article by William Pao on asmag.com, 19 July 2016.