A Study on the Dynamic Spatial-Temporal Trajectory Features of Pedestrian Small Group

Pedestrian small group is an important part of crowds in public places, they maintain similar speed and direction to reach the destination together through interactive cooperation. However, the research into the influence of groups on pedestrian flow dynamics has been limited. Previous research into the group behaviors mainly focus on the influence of group size on walking speed and spatial pattern problem from a static point of view. The dynamic spatial-temporal trajectory features of pedestrian small groups among people can provide significant insight into the group dynamics. This paper's aim is to provide quantitative insights into the dynamic spatial-temporal trajectory features of pedestrian small groups based on the real-world trajectories data including the interaction between pedestrians. These results will be a quantitative guidance for the group simulation model, the understanding of the social behavior in the video surveillance and automatic group identifying algorithm.