How streetscape shapes affect visitor emotions: An experimental analysis based on large-scale street view images in Xi'an, China

Abstract

The urban streetscape could generate a significant effect on individual emotions. Previous studies have predominantly utilized street view images to explore the emotional perceptions of residents, with less attention given to the emotions of visitors. This study aims to quantify the impact of shape characteristics within the visual features of urban streetscapes on visitors' emotions using large-scale street view images. Taking Xi'an, China, as a case study, we first constructed a dataset of visitors' emotional responses. Concurrently, we quantified the complexity and coordination of street shapes based on the outlines of street view images, serving as quantitative indicators of street shape. Subsequently, a random forest (RF) model was employed to fit the relationship between street shape characteristics and visitors' emotions. The results indicate that: (1) Visitors' six emotional experiences exhibit distinct spatial distribution patterns, and there is a certain homogeneity in the spatial distribution of emotions between visitors in non-habitual environments and residents in habitual environments. (2) The complexity and coordination of streets have a complex nonlinear effect on visitors' emotions. (3) Generally, the more pronounced the street shape characteristics, the more positive the emotional experience for visitors. When the shape complexity is 0.770 and the shape coordination is 1.555, visitors can achieve the optimal emotional experience across six dimensions. These findings can assist city managers in understanding visitors' visual preferences for streetscapes and provide guidance for optimizing street design.