In our system project, we want to investigate the question of which current methods of graphical data processing and which hardware can be used to automatically recognise the dynamics of a mass panic. The question is motivated by recurring catastrophes that are caused by the dynamics of a crowd and by people as a result of external events or by themselves. Automatic detection would allow targeted behavioural instructions to be given that could save lives. To this end, we want to investigate various questions in project form:

• How can individual people and their movements be automatically recognised in a crowd?
  
  There are various well-known algorithms for processing images that can be practically evaluated.
  
  
• Is it possible to realise recognition in which individual people can no longer be distinguished?
  
  Difficulties with recognition can arise due to greater distances, but also due to an oblique viewing angle or even a vertical view from above.
  
• What hardware is suitable for making detection possible?
  
  Camera images now offer high resolutions at low prices. Depending on the distance from the crowd and the number of cameras, image data of varying quality can be generated for further processing. Thermal imaging cameras can also be used as an alternative to visible light imaging. There are now also radar sensors that can detect movements at closer distances. A combination of different hardware could deliver better results.
  
• How reliably can a mass panic be recognised using the approaches known today?
  
  Recognising people and their movements can provide the basis. However, factors such as the distance between people and external events also play a role in the development of panic.
  

Depending on the number of participants, several questions and approaches can be investigated by smaller teams.

Contact

Prof. Dr Jens-Peter Akelbein and Prof. Dr Elke Hergenröther

Group 1

The aim was to create software that simulates a crowd of people in the form of corresponding movement and vital data.

The aim was to create software to warn of mass panics.

As part of the project, this project group dealt with the question of the extent to which a combination of regular location determination and infrared is a useful method for determining the density of people and for the early detection of a mass panic.