Modeling block-scale urban vitality: integrating streetscape perception and service accessibility

Urban vitality is a fundamental issue in urban planning and management research and a primary indicator for assessing the quality of urban development. Analyzing its spatial disparities at the block scale is essential for promoting urban sustainability. Residents’ perceptions of the built street environment significantly influence urban vitality. However, most existing approaches rely on macro-level indicators or linear assumptions, overlooking block-scale perceptual experiences and spatial heterogeneity, while also facing limitations in perception quantification and spatial autocorrelation. To tackle this, we integrate street view images and points of interest (POI) data to create a framework for vitality assessment that incorporates visual perception, environmental perception, and accessibility to service facilities. We utilize random forest and spatial error models to investigate nonlinear relationships between multi-dimensional indicators and urban vitality. The framework is implemented in the high-density central districts of Xi’an. Results indicate: (1) Urban vitality scores generally decline from the center to the periphery, with the highest value (0.63) found along the city’s central development axis. (2) The key variables’ contributions are ranked as lively score (0.31), commercial facilities density (0.26), and sky visibility index (0.17). (3) Spatial error regression reveals that the green visibility index (β = 0.23), commercial facility density (β = 0.24), and lively (β = 0.19) are strongly positively correlated with vitality, while the sky view index (β = −0.12) shows a negative correlation. These findings provide a comprehensive understanding of vitality distribution and key influencing indicators in high-density urban areas, offering new insights for enhancing vitality in peripheral zones and promoting urban sustainability.