The process of manually mapping pedestrian path networks for OpenStreetMap is difficult due to the large amount of data required and the potential for error. We developed a tool leveraging and integrating different globally-available data types for proactive pedestrian path and network data generation and traversal analytics incorporating accessibility. Along with existing street network data, state-of-the-art computer vision techniques are employed to automatically infer massive pedestrian path network information for 6 cities on 2 continents. The inferred pedestrian path network data are represented in OSM JSON format and can be directly used in OpenStreetMap as a data layer, or used in downstream routing and analytic applications.
No recording is available from this session.