Slide One

Line Crossing

Recognition of transit direction

Line-Crossing algorithms allow virtual lines to be drawn in scenes framed by video surveillance cameras, receiving automatic notifications if they are crossed by one or more specific objects.

Recognizing the transit of vehicles and people with MagicBox’s Artificial Intelligence provides extremely useful data and statistics that increase the safety of assets and people, both in private environments and in the public road system.

Use cases

PHYSICAL SECURITY

The ability to draw virtual perimeters enables MagicBox to automatically detect intrusions and unauthorized accesses, promptly alerting Control Rooms.

In these cases, intelligent rules can be set up to trigger numerous countermeasures – such as automatically initiating recordings for evidentiary purposes.

line-crossing-3

PEOPLE SAFETY

Virtual perimeters can also be used to monitor workplace safety, allowing to verify the presence of the minimum/maximum number of people in a specific area or the compliance of safety distances.

Being able to automatically monitor traffic directions also makes it possible to verify the respect of manoeuvring corridors, ensuring the safe transit of heavy vehicles.

ROAD SAFETY

Perimeters and virtual lines also have numerous applications in road safety, where they enable the identification of traffic direction violations and dangerous crossings.

Violations are automatically collected and archived, allowing for more in-depth investigations, the export of reports and the processing of appropriate statistics.

The universal AI video analytics system

With Line-Crossing algorithms, the monitoring of sensitive environments is optimized by allowing constant account of transits and accesses in a specific area. It is possible to detect:

  • accesses to restricted areas
  • intrusions and trespassing
  • non-compliance with traffic directions
  • transits in anomalous directions

The system also allows to count the number of events for statistical reports.

Smart video surveillance