Visual engagement of streetscapes: eye-tracking and cluster analysis of design elements

Streetscapes play a pivotal role in shaping pedestrian experiences and supporting livable, sustainable urban environments. This study investigates how the visual design of Green and Complete Streets (GCS) relates to pedestrian visual engagement. By categorizing streetscape elements into function-based, process-based, context-based, and design-based groups, the research examines their relationship with pedestrian gaze behavior. Two contrasting streetscape types – a GCS-inspired street and a Non-GCS street – were analyzed using principal component analysis (PCA), cluster analysis, and webcam-based remote eye-tracking. PCA identified three principal components: Mobility and Orientation Features, Safety and Accessibility Features, and Contextual and Transitional Features. Cluster analysis showed separation of the two predefined streetscape types in the reduced component space, with the selected GCS and Non-GCS stimuli occupying distinct positions across the three components. Eye-tracking comparisons showed higher aggregated image-level visual-engagement metrics for the selected GCS stimuli than for the Non-GCS stimuli, particularly for process-based and context-based AOIs, as reflected in dwell time, fixation count, and revisit count. These findings highlight associations between GCS-related streetscape features and patterns of visual engagement and support their consideration in the evaluation and design of pedestrian-oriented street environments.