Monitoring Surveillance Systems at all times can be incredibly costly. Utilising motion detection can reduce these costs through the detection of motion and the triggering of an alarm to notify the user that there may be activity that warrants their attention. The inability of this technology to differentiate between a genuine alarm and a false alarm by identifying what has been detected meant that there was still an over-reliance on people and resources to monitor systems accurately.
Video Analytics (VA) were introduced to detect patterns of motion, so that event triggers could be more specific, such as area intrusion or line crossing. This helped to increase efficiency, accuracy and customization, whilst reducing false alarms.
Deep Learning Video Analytics (DLVA) takes this to the next level, dramatically improving the level of accuracy of Analytics with a great impact upon false alarm filtering capabilities. However, costs associated with DLVA have limited its usage in the mass market as specialist recorders have been required to handle the processing needs of the technology.
Following several years of consultancy, research and in-house development, Videcon is delighted to announce the release of Concept Pro DLVA. Concept Pro DVLA is available on the latest ranges of Concept Pro Professional and Concept Pro Elite, Network Video Recorders. There is also a suite of traditional VA available on all Concept Pro NVRs. Through these developments, and a dramatic reduction in the cost of this technology, Videcon has ensured that all installers can take advantage of DLVA and VA technology in every application.
Traditional Video Analytics
Traditional VA were introduced to make surveillance more efficient and effective. Whilst VA has brought the industry forward, there remains doubt around its reliability. Indeed, many monitoring stations will not rely totally upon VA alarms and still require motion detection equipment, such as PIRs, to be the prime alarm trigger.
Videcon has implemented measures to improve the accuracy of traditional VA. This includes introducing a calibration wizard for quick set-up which applies depth to camera scenes by calculating the angle and height of the camera. The wizard then allows the user to configure whether an alarm is triggered by the centre point of a moving object or its ground point. For example, with an intrusion, it is appropriate to trigger an alarm when the centre of a moving object enters the detection area, but for a line crossing event the ground point is more accurate.
Videcon has made a suite of market leading VA available on all new Concept Pro recorders, with the set-up tools and configuration options in place to reduce false alarms and simplify deployment.
Deep Learning Video Analytics
Deep Learning Video Analytics (DLVA) radically advances false alarm filtering resulting in a CCTV system that can be relied on to accurately identify people, vehicles or animals.
DLVA provides a far more accurate approach to filtering out false alarms than motion detection and traditional video analytics. Instead of simply looking for changes in pixel values, DLVA utilises multi-layer, neural network algorithms to effectively learn the objects profile. It then uses a combination of pixel values, edges, vector shapes and a host of other visual elements to recognise the object as a person, vehicle or animal. The result is a system that will alarm only when an object of interest is in view, with vastly superior accuracy than traditional VA.
Following years of consultancy and R&D, Videcon has developed DLVA that is embedded in its range of Concept Pro Professional and Concept Pro Elite, NVRs, in conjunction with selected Concept Pro cameras.
Deep Learning Video Analytics and includes access to the following rules through the activation of a license key:
DLVA is extremely simple to set up and is done so by drawing an area of interest and selecting which rule to apply.
In the live user interface, a special DLVA tab on the timeline allows thumbnails of all
detected objects to be displayed for quick reference. Clicking on a thumbnail will instantly begin
playback of the original video. In addition to this, a new DLVA search mode allows rules to be applied retrospectively to recorded video to quickly identify when objects of interest were in view.
Videcon has introduced highly accurate, DLVA to the market on its cost-effective, Professional and Elite, Concept Pro Recorders. It has also improved VA reliability across its entire Concept Pro, Network Video Recorder range. This ensures that installers are no longer priced out of high-end, accurate, analytics applications where false alarms must be reduced and efficient monitoring is paramount.
To learn more, or to organise a demonstration of deep learning video analytics (DLVA) or traditional video analytics (VA), please call Videcon on +44(0)1924 528000.