Pedestrian–Environment Interactions: Multivariate Data-Based Correlations between Street Morphology and Urban Vitality

Urban vitality serves as a critical indicator of sustainable urban development. Its interaction with the built environment provides substantial support for future interventions. Using Yangzhou City as a case study, this paper investigated the correlations through morphological elements of streets at meso–macro scales in old city centers and newly built urban districts that influence urban vitality. Street units were delineated based on field surveys of pedestrian behavior patterns. By using deep learning algorithms and space syntax, street morphological features were extracted and mapped to the analysis unit together with street land use. Multiscale geographically weighted regression was utilized to determine the correlation between street vitality and its morphological elements. The results showed that pedestrian accessibility and facility street land use enhanced the vitality of urban streets in general, while the impact of retail and restaurant street land use on urban vitality presented a negative correlation trend. In addition, landscape shading in old city centers positively contributed to urban vitality. These findings were corroborated by photo records taken by residents and tourists at various time periods. This study has demonstrated that subtle modifications in urban form can influence urban vitality. Urban planning and governance strategies should place greater emphasis on regional contexts, as well as human behaviors and perspectives.