Deep Learning-based Identification of Urban Pedestrian Shade Facilities and Thermal Environment Assessment using Street View Imagery

Rapid urbanization and global climate change have intensified the urban heat island (UHI) effect, posing significant risks to public health and urban livability. Street-level shading is a vital passive cooling strategy to mitigate heat stress and enhance pedestrian thermal comfort. However, traditional methods often struggle to achieve large-scale, high-precision identification of diverse shade facilities—such as building overhangs, street trees, and specific structures like arcades (Qilou)—across the urban pedestrian network.

This research proposes a comprehensive framework leveraging Street View Imagery (SVI) and advanced AI analytics to bridge this gap. Initially, the study employs the YOLOv11 (You Only Look Once) deep learning architecture to automatically detect and quantify heterogeneous shading elements. By training on high-resolution SVI datasets, the model identifies multi-type shade facilities including building facades, arcades, street trees, and artificial awnings in complex urban environments.

Subsequently, the research evaluates the synergistic effects of these facilities on the pedestrian thermal environment. The extracted geospatial shade data are integrated with microclimate simulation tools to quantify their impact on thermal comfort indicators. Key parameters, such as the Sky View Factor (SVF), are derived from the pedestrian perspective to evaluate shading performance and its role in reducing heat exposure.

The findings are visualized through high-resolution thematic maps depicting shade coverage density and thermal comfort assessment results. This research provides urban planners and managers with scientific decision-making evidence to identify shade-deficient areas and optimize street designs for heat-risk reduction. By combining YOLOv11-based object detection with geospatial analytics, this study offers a scalable approach to enhance urban climate resilience and support sustainable, walkable urban development.