Analyse de réseaux urbains. Centralité, sinuosité et détection de raccourcis
Urban network analysis. Centrality, sinuosity and shortcut detection
This paper analyses the structure of roads (directed) and the pedestrian (undirected) networks in Lausanne (Switzerland). Centrality and sinuosity indexes are computed to perform spatial analysis at various scales. The study shows that the sinuosity index varies a lot at local (pedestrian) scale. The link between these variations and the underlying street patterns is investigated. The method also suggests that the comparison between Euclidean and shortest path distances can be used to improve pedestrian network by creating promising shortcuts, or alternatively to detect major network topological errors.
network analysis, spatial analysis, pedestrian network, centrality, sinuosity