Contextual analysis of pedestrian mobility in transport terminals

Abstract

Data acquisition using security cameras for pedestrian tracking and counting is inherently inaccurate due to limited video quality and the dynamics of pedestrian mobility. In this article we explore how specific patterns of pedestrian behavior in transport terminals can improve accuracy in tracking and counting. We show the effectiveness of our proposed techniques using data from the New York Grand Central Station. The proposed tool improves the accuracy of pedestrian labeling based on information about the dynamics of pedestrian mobility in the specific context of transport terminals.

Keywords

• Contextual analysis
• Pedestrian behavior
• Transport terminal monitoring