I based my thesis on the development of a database system that is capable of processing useful information from the incoming data via pedestrian's mobile devices thus providing usfeul information for security organizations deailng with crowd surveillance. These days news about tragedies in different crowded events are really common. Despite the fact that there are several systems intended to provide the aformentioned task, I believe that these cannot characterize effectively the status and dispersion of a larger crowd in close real time.
Unfortunately the models according to scientific literatures cannot be used directly in surveillance systems because they were mainly designed for simulations. For a successive solution I have to create a model (implemented as database as well), using currently existing models, that can provide an easy-to-use access to process the incoming data from any event.
I would also like to create a supplementary web interface where the organisers of said event can easly access the database using maps and diagrams.