Frontiers in Public Health, Journal Year: 2024, Volume and Issue: 12

Published: Sept. 18, 2024

Introduction Urban green space (GS) exposure is recognized as a nature-based strategy for addressing urban challenges. However, the stress relieving effects and mechanisms of GS are yet to be fully explored. The development machine learning street view images offers method large-scale measurement precise empirical analysis. Methods This study focuses on central area Shanghai, examining complex psychological perception. By constructing multidimensional perception scale integrating algorithms with extensive data, we successfully developed framework measuring Using scores from provided by volunteers labeled predicted in Shanghai's through Support Vector Machine (SVM) algorithm. Additionally, this employed interpretable model eXtreme Gradient Boosting (XGBoost) algorithm reveal nonlinear relationship between residents' stress. Results indicate that Shanghai generally low, significant spatial heterogeneity. has positive impact reducing effect threshold; when exceeds 0.35, its gradually diminishes. Discussion We recommend combining threshold identify spaces, thereby guiding strategies enhancing GS. research not only demonstrates mitigating but also emphasizes importance considering “dose-effect” it planning construction. Based open-source methods have potential applied different environments, thus providing more comprehensive support future planning.

Language: Английский