Visualizing congestion at large-scale events with an interactive-view system incorporating proximity-based networks

Abstract

Contact with infected individuals can lead to the spread of infectious diseases. During the COVID-19 pandemic, people were strongly urged to avoid the three Cs: closed spaces, crowded places, and close-contact settings. To hold large-scale events under such circumstances, reducing crowd congestion is key to preventing the further spread of infection. Therefore, identifying the pedestrian behaviors and walking patterns that pose a high risk of infection and utilizing them for effective crowd control is necessary. In this study, we propose an approach for visualizing walking paths while maintaining visibility from large-scale human flow data and representing both spatial and temporal features. The proposed method enables the visualization of the pedestrian proximity status as a network containing three components: a proximity network, proximity path, and pedestrian statistics that interact with each other. By operating the three components of this system interactively, we can observe the spatial and temporal features of situations with a high risk of infection during crowd congestion. An example of the operation of this system is presented by visualizing real-world human flow data measured at an event venue and identifying the proximity of the pedestrians.