基于开放数据的街道近人空间量化研究——以上海市为例

With the transformation of transportation and construction technology brought by the industrial development，the street space and building scale is very different to traditional cities in all respects，so classic urban design theories and control indexes which are applied to today’s street space already have some limitations. In addition，the widespread use of open urban data has brought new approaches and perspectives to street research. So the paper explores the scale of space at human with open data which could be used to provide an effective means to meet the psychological、visual and cognitive needs of small-scale under today’s large-scale、vehicle scale’s urban design..The paper introduces Street image, point of interest data and image recognition technology to get the important psychological impact factors and the indexes which impact people’s spatial perception and explores the factors and quantitative research on the spatial scale at human based on the indexes of western urban design, referring to the current situation of Shanghai city. Combined with other factors that affect the urban development and macro city form on the basis above with open street map and GIS，the paper presents proper indexes about the scale of space at human. Finally in light of street design guidelines overseas，the paper apply the indexes in Shanghai urban design guideline and concludes guidelines and suggestions towards the design of shanghai’s street space，meanwhile providing the statistic support for the planning regulations.

随着工业发展带来的交通与建筑技术的变革，街道空间与建筑尺度相比传统城市已发生巨大变化，经典城市设计理论应用于当今街道空间控制已显示出一定局限性。此外，城市开放数据的广泛应用为街道研究带来了全新的途径和视角。由此，本课题引入近人空间视角并结合开放数据探讨在当今大尺度、车尺度的背景下，满足人的心理、视觉、认知需求的小尺度街道设计。.研究立足上海城市现状，利用街景图像、兴趣点数据与图像识别技术对近人空间的客体指标与个体心理感知展开探索，构建数理模型来辨别与街道感知相关的关键因子及数值，从而将人的复杂心理感知进行定量化研究。在此基础上，结合开放地图数据与GIS技术探索近人空间与城市建设因素、宏观城市形态之间的联系，得到适宜的量化指标。最后借鉴国外街道设计导则，将量化指标应用于上海相关的街道设计导则，为制定新的街道设计标准提出建议与依据，也为规划法规的修订提供数据支撑。

随着工业发展带来的交通与建筑技术的变革，街道空间与建筑尺度相比传统城市已发生巨大变化，经典城市设计理论应用于当今街道空间控制已显示出一定局限性。此外，城市开放数据的广泛应用为街道研究带来了全新的途径和视角。由此，本课题引入近人空间视角并结合开放数据探讨在当今大尺度、车尺度的背景下，满足人的心理、视觉、认知需求的小尺度街道设计。.课题选取上海市内环内的街道为研究案例，利用多源数据与图像分割、智能识别技术对街道的物理特征、界面特征、城市功能、宏观形态四类指标及个体对街道的宜人性与安全性感知展开探索，构建数理模型来辨别与街道感知相关的关键因子及数值，对上海街道设计提出指引与建议，也为规划法规的修订提供数据支撑。研究发现：绿视率是影响街道宜人性的关键因子；建筑密度、高宽比对于宜人性感知具有积极影响。城市功能是影响街道安全感的关键因子，其中街道功能密度比功能混合度对安全感的感知影响更大；店招个数、绿视率对于安全感知也具有积极影响。