人本尺度的街道空间品质测度——结合街景数据和新分析技术的大规模、高精度评价框架

Human-scale Quality on Streets: A Large-scale and Efficient Analytical Approach Based on Street View Images and New Urban Analytical Tools

本研究针对城市微更新的实际需求,结合街景数据和新分析技术提出了面向人本尺度的街道空间品质测度操作框架。研究以上海杨浦区和虹口区为案例,基于街景图像数据,运用机器学习算法对街道空间要素进行提取,进而使用神经网络算法训练评价模型,构建大规模且精细度高的街道场所品质测度。与此同时,通过叠加sDNA的空间网络可达性分析结果,建立以"品质评价"与"可达性分析"为维度的评价矩阵,找出分析区域中"具有更新潜力的街道",为城市微更新提供精细化技术支持。

This study provides an operational framework about street quality measurement by the means of large-scale data analysis at the humanistic scale and the results can be regarded as the benchmark for the renewal of urban street space. Taking Hongkou District and Yangpu District of Shanghai as an example, based on Street View Images (SVI) data, this paper takes advantage of machine learning to extract spatial feature, then uses neural network (ANN) to measure the quality of street places with wide distribution and fine resolution. Besides that, an evaluation matrix established by overlapping analysis will combine quality evaluation with network accessibility analysis (sDNA). Finally, we find out those “potential streets” and provide fine theoretical foundation for urban micro-renewal.