ICT4SS - ICT FOR SMART SOCIETIES
Automatic passengers counting systems in local public transport through computer vision and video analysis techniques
Thesis in external company
Reference persons ENRICO MAGLI
External reference persons Alberto Forchino, Rossella Trancuccio (GTT)
Thesis type APPLIED RESEARCH
Description The project will be carried out at the GTT company and consists in the evaluation of automatic passengers counting systems in local public transport. Particularly, the video surveillance already existing on vehicles (not dedicated to passenger counting) will be used and video analysis techniques will be applied, in order to quantify the number of passengers boarding/alighting per each stop and, therefore, obtain the number of people on board.
The aim of the project is to develop video analysis techniques and compare the data obtained with a manual passenger count, in order to define which accuracy is achievable.
Next, you should combine the information obtained from this count with those obtained from a second system, based on the pressure load of the suspensions (which is already occupying a student in Civil Engineering), acquiring them in a db from the packets flowing on the local bus network.
Required skills Computer vision and video analysis techniques, knowledge of libraries like Opencv or similar, programming of Single Board Computer (eg Raspberry PI).
Deadline 12/04/2019 PROPONI LA TUA CANDIDATURA