Measuring pedestrian-level street greenery visibility through space syntax and crowdsourced imagery: A case study in London, UK

Abstract

Linear green spaces, composed of street trees, shrubs, and grass, provide diverse opportunities for human-nature interaction. However, current research tends to visualize street greenery from a single perspective, such as images or planar analysis and neglects pedestrian-scale street tree visual analytics. Space syntax relies on precise urban context data input, and street view image analysis does not cover sidewalk greenery visibility. This study integrates green visibility analysis based on space syntax's Visibility Graph Analysis (VGA) with the Pedestrian Green View Index (PGVI) calculated from street imagery to propose a comprehensive evaluation model for pedestrian-scale green visibility. To validate the accuracy of the methods, we established a participation scoring system involving 183 volunteers to collect their green perceptions of nine types of streets in the City of London, UK. The study reveals a complex relationship between VGA and PGVI, with VGA providing a robust, geometric-based visibility measure and PGVI offering a qualitative, human-centric perspective on urban greenery. Our findings indicate a significant correlation between PGVI and human evaluations, affirming PGVI's potential to reflect pedestrian experiences, while highlighting the limitations of VGA in capturing the nuanced, multi-dimensional aspects of human perception. This underscores the necessity of integrating human feedback in urban planning tools to ensure a comprehensive understanding of green spaces. Future research should enhance methodological rigor by incorporating temporal and seasonal dynamics, expanding datasets, and exploring the interplay between green visibility and other environmental factors.