The main campus of Swinburne University of Technology (SUT) is located in the inner-suburb of Hawthorn.
This presents various challenges for the security team. To overcome these challenges, SUT has invested in a state-of-the-art Milestone CCTV surveillance system with over 650 Axis cameras to assess situations and assist in the protection of its students, staff, and building facilities across its campuses - especially after hours.
Swinburne have excellent CCTV surveillance infrastructure in place with great penetration and coverage across their campuses, but with hundreds of cameras, there is invariably more video-data streaming into the command-center than the security team can easily “live” monitor. They were looking for new technologies that could help filter the feeds and present only those incidents requiring closer inspection.
In searching for a solution a key issue was that the analytics had to perform in a very busy 24x7 environment and didn’t require complex programming and ongoing maintenance. A lot of video analytics companies over the last 5-10 years have tried to tackle the problem of automating surveillance. Whilst the systems perform well in certain scenarios, they are challenged when the number of cameras and scene complexity increases.
iCetana's system was designed to use a different approach specifically to solve these problems. The system automatically learns and detects “all kinds” of events without the customer having to program anything.
The software was installed to monitor known hotspots across all five Melbourne campuses. As iCetana is a Milestone Solution Partner, the system was designed to easily integrate into Swinburne’s Milestone VMS environment. iCetana's LiveWall plugin showed any detected events on a single LCD screen.
By focusing on events displayed on the LiveWall screen operators were able to easily monitor large numbers of cameras with a single view rather than scanning banks of video monitors. Typically the events detected and shown on the LiveWall represented only 1% of the total video footage streaming from the camera network to the control room – effectively reducing the operator’s load and enabling them to focus on other important tasks. Of the incidents detected many were the typical security events such as vandalism, loitering, fights, anti-social behaviour and vehicle violations.