Street view versus remote sensing greenery – comparison of two exposure metrics across urban-rural settings

Urban greenery, as a critical urban landscape component, plays an important role in improving the living environments’ and residents’ well-being. Previous studies have predominantly adopted satellite image-based vegetation measurements. This study aims to quantify pedestrian-perspective greenery visibility using Google Street View (GSV) images and to understand how greenery types and built environment characteristics influence the correlation between pedestrian and aerial greenery assessments. We collected GSV images located on 34,601 sampling points and applied the DeepLab v3+ deep learning model to quantify green view index (GVI) from the pedestrian perspective. We distinguished green vegetation view index (GVVI) and green terrain view index (GTVI) to differentiate vertical and horizontal greenery types. Normalized difference vegetation index (NDVI) was extracted from Sentinel-2 images using circular buffers of varying radii (10–200 m) centered on GSV sampling points. Sampling points were filtered based on the buffer distance to avoid overlapping NDVI pixels in neighboring sampling points. Spearman correlation analysis was conducted across different typologies (urban, intermediate, rural) to examine GVI-NDVI relationships. Street-level greenery exhibited substantial spatial heterogeneity across the whole of the study area (Basel, Switzerland). GVI vs. NDVI in buffers with different radii had strong positive correlations, with a maximum Spearman coefficient of 0.77 for the 15 m NDVI buffer. Correlation coefficients decreased progressively from urban (0.77) to intermediate (0.72) and rural (0.66) areas. Correlation coefficients strongly decreased with increasing buffer sizes. Analysis of GSV images with high NDVI but low GVI values indicates that greenery types and building distributions significantly affect the street-level visible greenery. This study links street-level greenery with features in the built environment by using different methods for assessing green exposure. The findings provide methodological insights for greenery exposure studies and inform evidence-based urban planning strategies for optimizing green visibility.