Integrating Multi-Source Data to Construct Complete Pedestrian Networks

Abstract Accurate pedestrian network modelling is essential for pedestrian dynamics. However, existing pedestrian network datasets are fragmented, lack connectivity and primarily focus on vehicle-oriented infrastructure, which limits their effectiveness for fine-scale pedestrian applications. This study presents a graph-based model that is designed using multiple spatial datasets to enhance pedestrian network construction. This framework is applied in a case study in Taipei City, which produces a promising preliminary pedestrian network compared to the existing pedestrian datasets.