Quantification of the impact of street design features on restorative quality in urban settings

This paper investigates the impact of design of urban spaces on restorativeness. It aims to identify the urban street design features that are highly effective in shaping human restorativeness and quantify their impact on restorativeness of urban dwellers. The study employs a suite of data acquisition methods, including crowdsourcing, computer vision (CV), and Geographic Information Systems (GIS), to gather data on people's perceptions of urban environments that feature different configurations of urban street elements. Machine learning was used to identify the influential urban street design elements on human restorativeness and quantify impacts. Our findings reveal that while the amount of greenery generally enhances restorativeness along with sky visibility, an excess beyond a certain threshold diminishes its positive effects- hence indicating a strong non-linear relationship between sky visibility and greenery density in relation to restorativeness impact of such urban spaces. This suggests that a balance of greenery is essential for promoting restorativeness in urban environments. Results also indicate that height-of-buildings, irregular-building-height, building-density, crowdedness, and retail-stores are negatively associated with restorativeness while around urban spaces. Practitioners can benefit from these findings as this study provides one of the comprehensive computational evaluations of urban street design elements towards people's restorativeness in urban settings.