Fraunhofer Institute of Optronics, System Technologies and Image Exploitation IOSB 
The Challenge: Security in real time
Major events, train stations, airports, industrial facilities – anywhere large numbers of people gather in confined spaces, dangers can arise quickly and without warning. Critical crowding, sudden changes in crowd flow, or bottlenecks often develop within seconds and require immediate, coordinated action.
However, the reality of many current operational situations shows that event organizers, security services, and authorities such as the police and fire department often work with isolated information. This fragmented information base makes it difficult to maintain the necessary overall overview in an emergency – precisely when rapid, interagency coordination is most urgently needed. Furthermore, crowds are highly dynamic. Conventional, analog observation methods are barely capable of capturing these complex developments in a timely manner and evaluating them thoroughly.
Our Approach: An objective situation picture instead of vague estimates
GeoVID is an AI-supported system developed at Fraunhofer IOSB for video-based detection and analysis of crowds (crowd monitoring) in distributed urban environments. The focus is not on the individual, but on the collective behavior of the crowd: crowd density, distribution, directions of movement, and speeds are captured in real time and displayed on a shared, interactive situational picture. This creates an objective, shared situational picture for all stakeholders – without delay and without room for interpretation.
