The algorithm allows to detect the aggregation density of people in a video surveillance area.
The algorithm can be integrated into a system that allows the formation of a crowd or the presence of groups to be reported in real-time.
The crowd detection node sends an alert to the platform, that is to the command and control system, when it detects an increase in the density of people aggregation.
The operator of the command and control system:

• receives the alert of this type coming, for example, from a camera that monitors an open area inside the park, or on the streets;
• processes the alert by viewing the images received on the platform;
• if necessary it proceeds with a follow-up: it could decide to directly access the streaming of the context camera (using the video surveillance management software), to make sure that the group of people can present a real threat or not.

The solution is plug & play and easily integrates with the pre-existing infrastructure.

Implementation

Ideal for monitoring the density of aggregation within means of transport (buses, trains and metro) and within a delimited space (congresses, events, squares) for the purposes of safety and prevention and in addition as a containment measure of Covid-19 infection

Background

ATM - Azienda Trasporti Milanese