随着城镇化的加速， 城市环境中的拥堵、噪音、信息过载等问题逐渐凸显，人们更加容易遭受焦虑抑郁等心理问题。不良的城市环境威胁着城市居民的身心健康，WHO报告指出，环境风险每年至少导致 1300 万人死亡，约占全球死亡人数的 24%，健康的环境对人类健康和发展至关重要。

社区为城镇居民提供了居住和生活的空间，大量研究证实了社区环境具有改善居民情绪、恢复精神压力和降低生理应激的恢复性效益，对促进居民的生理健康和心理健康有重要意义。然而，目前国内对社区环境恢复性效益的研究成果集中于自然要素，对城市环境中占比较大、更易于控制与改善的人工要素缺乏讨论。

本文着眼于社区视觉环境的恢复性效益，提出了一种全面客观量化社区视觉环境特征的方法和主客观结合的环境恢复效益评价方法，使用图像语义分割和可视域分析（ Isovist Analysis）提取了社区视觉环境的特征要素，基于皮肤电导反应（ SCR）和心率变异性（ HRV）量化生理恢复效益。其次，全面分析了社区视觉环境特征要素对恢复性效益的影响，通过相关分析、空间聚类、多元 Logistic 回归，挖掘出了视域面积、遮掩度、游离距等空间特征和道路、天空、植物等物质要素对恢复性效益的显著影响关系，并根据分析结论提出了社区视觉环境建设与优化建议。此外，建立随机森林模型（R2>0.8） 进一步探讨了社区视觉环境要素与情绪恢复的潜在复杂机制，揭示了人类要素、道路要素、交通要素等平面视觉要素对情绪恢复水平的非线性关系。最后，采用非支配排序遗传算法II（NSGA-II）和逼近理想解排序方法（TOPSIS）得到了恢复性效益最佳的社区视觉环境特征要素推荐取值区间。研究的理论成果在真实的社区环境中得到验证，其准确性与泛用性得到了检验， 对社区恢复性环境建设有重要的理论指导意义，有助于推进健康中国发展战略的落实。

关键词：社区；视觉环境；恢复性效益；影响分析

As urbanization accelerates, problems such as congestion, noise, and information overload in urban environments are gradually becoming more prominent, which makes people more prone to suffer from psychological problems such as anxiety and depression. The poor urban environment threatens the physical and mental health of urban residents. WHO reports that environmental risks cause at least 13 million deaths each year, accounting for about 24% of global deaths, and that a healthy environment is vital to human health and development.
Communities provide urban residents with living and life spaces. Numerous research has confirmed the restorative benefits of community environments, which are of great importance in promoting residents' physical and mental health, including improving residents' moods, restoring mental stress, and reducing physiological stress. However, current research on the restorative benefits of community environments in China is focused mainly on natural elements, with little discussion of the artificial elements that make up a significant proportion of the urban environment and are more easily controllable and improvable.
This article focuses on the restorative benefits of the visual environment in communities, and proposes an objective and quantitative method for measuring the characteristics of community visual environment and a combined subjective and objective method to evaluate the restorative benefits of the environment. Image semantic segmentation and Isovist Analysis are used to extract the characteristic elements of community visual environment, and the physiological restorative benefits are quantified by skin conductance response (SCR) and heart rate variability (HRV), which complements subjective restoration benefits. Secondly, the influence of community visual environment characteristic elements on restorative benefits is comprehensively analyzed. Through correlation analysis, spatial clustering, and multiple logistic regression, the significant relationship of spatial characteristics such as Isovist area, occlusivity, drift magnitude, and material elements such as road, sky, plant on the restorative benefits are explored, and suggestions for community visual environment construction and optimization are proposed based on the analysis conclusions. In addition, random forest models (R2>0.8) are established to further explore the potential complex mechanism of community visual environment elements and emotional restoration, revealing the non-linear relationship between planar visual elements (such as human, road, and traffic) and the emotional restoration. Finally, the recommended value range of community visual environment characteristic elements with the best restorative benefits are obtained using the nondominated sorted genetic algorithm II (NSGA-II) and technique for order preference by similarity to an ideal solution (TOPSIS). The theoretical results of the study are validated in a real community setting, of which the accuracy and universality are tested, providing important theoretical guidance for the construction of community restorative environments and contributing to the implementation of the Healthy China Strategy.