Constructing child-friendly cities: Comprehensive evaluation of street-level child-friendliness using the method of empathy-based stories, street view images, and deep learning

Abstract

Despite the fact that streets are a crucial element in the development of child-friendly cities, existing street planning often overlooks child-friendliness. With the Gulou District of Nanjing as a case study, we used the method of empathy-based stories to assess the perception of children regarding the urban street environment and deep learning to evaluate the child-friendliness levels of streets on a large scale. Furthermore, we explored the effects of different street view elements on the child-friendliness and identified effective improvement strategies. The results showed that the streets in Gulou District generally had low child-friendliness levels, where low-level streets comprised 41.635 %, and extremely high-level streets only comprised 0.407 %. The spatial analysis revealed that areas with higher child-friendliness were concentrated in southeastern Gulou District, whereas areas with lower friendliness were predominantly on its periphery. Our work also indirectly reveals potential differences in the perception of street elements between children and adults. Children tend to favor artificial elements like buildings, walls, and fences. Our research emphasizes the urgency and importance of improving the existing street environment to support the development of child-friendly cities.