In English

Tracking Guest Flows at Night Clubs

Anton Bengtsson ; Ulrik Börjesson ; Oscar Evertsson ; Martin Fredriksson ; Simon Gustafsson ; Adam Wiklund
Göteborg : Chalmers tekniska högskola, 2017. 77 s.
[Examensarbete för kandidatexamen]

Creating new solutions or optimizing already existing ones has been in the human nature as far history can tell. Necessarily it does not always have to be the revolutionary, such as flying to the moon. It could be smaller things as optimizing the travel time by a couple percent. In businesses today, one of the most expensive costs are personnel. Therefore, optimizing this resource could be a good way for companies to further develop their businesses. Nightclub Data AB (VNU) makes digital clickers for tracking the flow of people at the entrances of night clubs. To further track the flow of people within the night clubs, rather than just the entrance, the company has commissioned a research regarding how this could be achieved. The goal of this project was therefore to develop a prototype able to count the amount of people inside a night club. Furthermore this data was supposed to be visualized in an intuitive way for night club managers and staff in order to optimize their business. A pre-study was carried out in order to explore and evaluate different possible methods and technologies. The methods and technologies that were found to be most suitable was then tested in the process of developing a prototype. The resulting prototype captures thermal images which are then processed by binarization and white pixel area calculations in order to count the amount of people in them. This data is then sent to a database that is accessed and visualized for the end-user in a web interface that integrates with VNU’s existing system. In conclusion a prototype was successfully developed for tracking the guest flows at nightclubs. The prototype fulfills the requirement specified early in the project with exception of cost. However, creating a complete system for tracking the guest flow was harder than expected and contains room for possible improvements.

Nyckelord: Camera, tracking, night clubs, Matlab, image analysis, human movement.

Publikationen registrerades 2017-09-07. Den ändrades senast 2017-09-07

CPL ID: 251745

Detta är en tjänst från Chalmers bibliotek