Data and code

Aligning with the United Nations’ Sustainable Development Goals, the focus on creating safe, sustainable cities and enhancing the wellbeing of individuals across all age groups has become a central aspect of urban planning and environmental management. The environments we live in significantly influence our thoughts, emotions, and interactions with the world around us. Therefore, it is vital to explore how people perceive their surroundings and how these perceptions differ among various social groups, particularly in relation to social inequities in environmental exposure. Additionally, there is an ongoing need to address how improvements in the physical environment can enhance wellbeing and reduce these disparities within urban spaces. As such, our study aims to unveil the social inequity of neighborhood visual environment across different social/vulnerable groups (i.e., White, Black, Asian, Hispanic, low-income, low-educated, and unemployed) via crowdsourced street view imageries and computer vision and further examine which built environmental features that are associated with people’s visual perception towards the surrounding environment via multi-model machine learning methods, with the pilot study in Los Angeles County. Neighborhoods with a high concentration of Black, Hispanic, and low-income, low-educated and unemployed populations have a higher level of boring and depressive perception while a lower level of beautiful, liveable, safe, and wealthy perception. The most important actual built environment features positively associated with the neighborhood soundness include the density of canopy, followed by the density of multiple units, the distance to CBD, and car commuting to destinations, regardless of social groups. The perceived visual environment in the neighborhoods with a high concentration of Black, Hispanic, and low-educated groups can be not well explained by actual built environment features, reflecting the possibility that other underlying variables (e.g., personal preference, place attachment or other subjective factors) confound their perception on surrounding environment. The conceptual framework and analytical workflow used in this study can be readily applied in cross-disciplinary studies more broadly, to guide through urban planning, urban design, and healthy city initiatives with place-based evidence.

契合联合国可持续发展目标（Sustainable Development Goals, SDGs），打造安全可持续城市、提升各年龄段人群福祉，已成为城市规划与环境治理的核心议题。我们所处的生活环境，对自身的思维、情绪及与周遭世界的互动均具有显著影响。因此，探究人们如何感知周遭环境，以及不同社会群体间的感知差异——尤其是与环境暴露相关的社会不公问题——具有重要意义。此外，学界仍需持续探索如何通过优化建成环境提升民众福祉，并缩小城市空间内的此类福祉差距。据此，本研究以洛杉矶县为试点区域，借助众包街景影像（crowdsourced street view imageries）与计算机视觉（computer vision）技术，揭示不同社会/弱势群体（即白人、黑人、亚裔、西班牙裔、低收入群体、低学历群体及失业群体）的邻里视觉环境社会不公问题；并通过多模型机器学习（multi-model machine learning）方法，探究与民众对周遭环境的视觉感知相关的建成环境特征（built environmental features）。黑人、西班牙裔、低收入、低学历及失业群体占比较高的邻里空间，其民众感知到的枯燥感与压抑感更强，而对美感、宜居性、安全性与富足感的评价则更低。无论所属社会群体为何，与邻里环境优良度呈显著正相关的核心建成环境特征依次为：树冠覆盖率、多户住宅密度、距中央商务区（Central Business District, CBD）的距离，以及驾车通勤出行占比。在黑人、西班牙裔及低学历群体占比较高的邻里空间中，民众感知到的视觉环境难以通过实际建成环境特征得到充分解释，这反映出其他潜在变量（如个人偏好、地方依恋（place attachment）或其他主观因素）可能会干扰他们对周遭环境的感知。本研究采用的概念框架与分析流程，可广泛应用于跨学科研究，为基于实证的城市规划、城市设计与健康城市建设实践提供指导。