Visualization system for human mobility analysis

摘要

City-scale events, such as festivals, sport events, concerts gather thousands of people in different areas of a city. By installing large area Wi-Fi routers which record the presence of people using devices such as smartphones and tables, the density and the movements of the crowds can be estimated and classified. Analysis of crowd data collected during such events leads to valuable insights and feedback both for the parties involved and for city planners and advertisers. In this paper we present a analysis and visualization system for improved understanding of density and movement information. We discuss real datasets constraints, data clean-up techniques and methods for estimating and classifying the crowd movements. These methods are implemented inside a framework that allows easy integration and analysis of mobility datasets. As proof of concept we present the analysis of a real dataset collected during a city-scale experiment in Arnhem, the Netherlands.