Heterogeneous Channels of Sky and Glazed Facades Affecting Street Perception: A Verification Based on Street View Imagery

Light exposure in urban street environments significantly shapes pedestrians’ subjective perception. While previous studies based on street view imagery (SVI) have predominantly regarded sky regions as the sole source of such exposure, glazed facades also contribute significantly through reflected light. Taking the Tianhe District of Guangzhou, China, as the study area, this research utilized luminance extraction and deep learning techniques to process SVIs and constructed an interaction regression model to analyze the channels through which light exposure affects pedestrian perception, as well as the specific role of glazed facades in this process. Key findings include: (1) elevated light exposure negatively affects overall perception, serving as a critical constraint on environmental quality; (2) under equivalent exposure conditions, glazed facades, although their overall main effect on perception remains negative, can partially buffer the perceptual burden induced by sky-dominated exposure; (3) based on these empirical channels, we propose classification and targeted renewal strategies for streets. This study proposes an integrated framework comprising “Sky/Glazed Facades—Light Exposure—Perception”, successfully incorporating reflected light from glazed facades into SVI-based luminous environment research. It reveals the differentiated channels of light exposure effects, providing guidance for both building forms and facade materials across urban blocks.