The Computational Image of the City

The repository contains a series of notebooks for extracting the Computational Image of the City from geospatial datasets

Introduction

This repository guides the user through the extraction of the Computational Image of the City; In other words, it allows to automatically identify a set of salient urban features in line with the definitions laid down by Kevin Lynch in The Image of The City using open and freely available geospatial datasets.

The analysis workflow is written in Python and relies on the library cityImage, and its dependencies (mainly: Geopandas, NetworkX and OSMNx)

The methods are fully documented in A Computational approach to ‘The Image of the City’ by Filomena, Verstegen, and Manley, published in Cities.

Notebooks' purposes

• 1a_Nodes_Paths_Districts_Web.ipynb It supports the identification of * Nodes, Paths* and Districts from street network's graphs automatically extracted from OSM. The user does not need to provide any kind of data.
• 1b_Nodes_Paths_Districts_Local_Files.ipynb It supports the identification of Nodes, Paths and Districts from the street network's graph obtained from a .shp file. The user is encouraged to use this notebook if they wish to work with a street network .shp file in their possession.
• 2a_Landmarks_Web.ipynb It identifies Landmarks from a set of buildings automatically extracted from OSM. As the height field may not be complete in OSM; some landmarkness components are not computed here.
• 2b_Landmarks_Local_Files.ipynb It identifies Landmarks from a set of buildings in a .shp file provided by the user. This is the recommended and complete way to extract landmarks (if the buildings' height field is complete).
• 3_Barriers.ipynb It supports the identification of Barriers (Edges in The Image of the City) from several features in OSM (roads, railways, waterbodies and waterbanks, parks).
• 4_Street-Network_Simplification.ipynb - This notebook presents a complex pipeline to automatically simplify a street network (either obtained from OSM or local files), as concerns: roundabouts, complex junctions, dual lines.