Building a robust perimeter security solution: Merging Radar and Camera

Magos is at the fore-front of technology, bringing a complete AI-based integrated radar and PTZ camera solution to the security market.

The complete “MASS+AI” solution, is composed of Magos radars, with detection ranges ranging from 250 meters up to 1000 meters, PTZ cameras with IR laser illumination, virtually of any brand, and a powerful Magos analytics appliance. The solution is able to merge the information from all the sensors, providing a comprehensive and robust situation awareness solution and thus achieves zero nuisance alarm and maximum sensitivity and detection probability.

MASS+AI solution

Magos has identified the weak spots of current technologies and built a solution that extracts the “best of both worlds”. Radars are very reliable movement detection sensors, that can reliably detect and track potential intruders and pin-point their exact world position at any moment. They have a wide field of view, enabling constant surveillance and concurrent tracking of multiple targets spread throughout a large perimeter.

However, radars have a hard time classifying targets – is it a person, a dog, a large bird, maybe it’s a line of trees moving together during a wind gust? This may cause a radar only detection solution to emit nuisance alarms caused by non-relevant targets. On the other hand, cameras usually have a narrow field of view, especially when they are required to provide a clear image of a target and are forced to zoom in on the target. While zoomed in, they miss the movements occurring around them in other areas of the perimeter.

Recent advances in AI based image classification provide amazing results in comprehending the image scene and distinguishing between different objects in the scene (human, animal, vehicle, bike, etc…). These advanced algorithms, however, require an even more clear and zoomed-in image of the object to reliably classify it.

Current video analytics engines usually rely on video motion detection and are very sensitive to lightning and weather issues like shadows, flares, cloud movement, rain, wind gusts, etc. This makes video-only based security solutions very hard to calibrate and usually results in a high number of false alarms. Furthermore, all these solutions require a static camera and cannot work with PTZ (Pan-Tilt-Zoom) cameras, as the analytics engine relies on the scene to be static.

Therefore, such solutions are very limited in their coverage and range, and eventually require a very large number of cameras to be installed along the perimeter (every 50 meters for example). The infrastructure costs of a solution like this are enormous – trenching, multiple pole installation, communications and power infrastructure, outdoor field boxes for concatenating each pole and installing the equipment itself on typically high poles using man-lifts. This renders the video-based perimeter security solution complex, costly, and also considerably increases the maintenance costs due to large number of sensors and infrastructure.

The alternative offered by Magos reduces the number of poles required by an order of roughly 1 to 20, relying on the large coverage of the radar. Static cameras are replaced by PTZ cameras that can provide target verification and nuissance alarm filtering. A single PTZ camera (coupled with a radar) can replace over 20 static cameras since the radar can automatically point the PTZ camera to the exact point of interest, and apply appropriate zoom level such that the target is clearly visible within the frame.

This greatly reduces total costs of perimeter security projects as the infrastructure needed is much “lighter”, less video channels are needed so the video management costs are also reduced, less physical sensor are installed so MTBF (Mean time between Failures) is higher and maintenance is simpler.

In other words, it’s a “no brainer” that the MASS+AI solution offers a superior solution in terms of detection performance at a fraction of the cost of traditional security technologies based solutions. 


Author: Aviel Kisliansky

    Acceda a la última información con nuestro boletín: