The Visual Greenery Field: Representing the Urban Green Visual Continuum with Street View Image Analysis

This study proposes a method to analyze urban greenery perceived from street-level viewpoints by combining geographic information systems (GIS) with image segmentation. GIS was utilized for a geospatial statistical analysis to examine anisotropy in the distribution of urban greenery and to spatialize image segmentation data. The result was the Visual Greenery Field (VGF) model, which offers a vector-based representation of greenery visibility and directionality in urban environments. The analysis employed street view images from selected geographic locations to calculate a Green View Index (GVI) and derive visual vectors. Validation confirmed the reliability of the methods, as evidenced by solid correlations between automatic and manual segmentations. The findings indicated that greenery visibility varies across the cardinal directions, highlighting that the GVI’s average value may obscure significant differences in greenery’s distribution. The VGF model complements the GVI by revealing directional coherence in urban greenery experiences. This study emphasizes that while the GVI provides an overall assessment, integrating the VGF model enriches the understanding of perceptions of urban greenery by capturing its complexities and nuances.