Search

Streets play a crucial role in the pedestrian catchment area (PCA) of metro stations. However, the large-scale quality measurement of street space and its influence on the vitality of station area have not been well revealed. With multisource big data such as points of interest (POI), and street view images, a three-dimensional evaluation system based on the pyramid scene parsing network (PSPNet) and spatial design network analysis (sDNA) is constructed. 73 metro stations in the Third Ring Road of Chengdu are chosen as research samples to carry out large-scale quantitative evaluation of street space in PCAs to reveal the quality characteristics of street space at the overall urban, PCA, and circle scales. Furthermore, this study constructs two multiple linear regression models of weekdays and weekends to explore the relationship between urban vitality and street space quality indicators. The results indicate a heterogeneous distribution of street quality on an urban scale. Streets located in the 300–500 m of PCAs rate highest in terms of convenience and the overall street space quality. The functionality dimension of street spaces in the sample PCAs of Chengdu present a gradient effect with the highest score of 0–300 m in the circle, while the comfortability dimension of streets shows an opposite trend. The multiple linear regression analysis show that street quality indicators are more explanatory of the weekday vitality than the weekend vitality. It indicates that well-connected street network, pleasant street scale, and abundant urban facilities have the greatest effect on urban vitality in the PCAs. The findings can provide new ideas for making targeted interventions in the urban design of metro station areas, to improve the quality of streets and foster urban vitality. [ABSTRACT FROM AUTHOR]