叶宇 张昭希 张啸虎 曾伟

Ye Yu, Zhang Zhaoxi, Zhang Xiaohu, Zeng Wei

关键词：空间品质；可达性；机器学习；人本视角；街景数据；街道

Keywords：Street Quality; Accessibility; Machine Learning; Human-centered Perspective; Street View Image; Street

摘要：

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。

Abstract：

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.

叶宇，同济大学建筑与城市规划学院助理教授，硕士生导师

张昭希（通信作者），同济大学建筑与城市规划学院硕士。zhangzhaoxi527@163.com

张啸虎，新加坡—麻省理工联合研究中心（SMART）博士后

曾伟，中科院深圳先进技术研究院副研究员

• [1] 龙瀛, 叶宇. 人本尺度城市形态: 测度、效应评估及规划设计响应[J]. 南方建筑, 2016(5): 41-47.

  [2] JACOBS J. The life and death of great American cities[M]. New York: Random House, 1961.

  [3] LEFEBVRE H. ‘Notes on the new town’, introduction to modernity[M].London: Verlso, 1962.

  [4] GEHL J. Life between buildings: using public space[M]. New York: Island Press, 1971.

  [5] WHYTE W H. The social life of small urban spaces[M]. The US: The Conservation Foundation, 1980.

  [6] LYNCH K. Good city form[M]. Cambridge: The MIT Press, 1984.

  [7] 龙瀛, 沈尧. 数据增强设计——新数据环境下的规划设计回应与改变[J]. 上海城市规划, 2015(2): 81-87.

  [8] LIU X, SONG Y, WU K, et al. Understanding urban China with open data[J]. Cities, 2015(47): 53-61.

  [9] TORII A, HAVLENA M, PAJDLA T. From google street view to 3d city models[C]. ICCV Workshops, 2009: 2188-2195.

  [10] NAIK N, PHILIPOOM J, RASKAR R, et al. Streetscore-predicting the perceived safety of one million streetscapes[C]. 2014 IEEE Conference on Computer Vision and Pattern Recognition Workshops, 2014: 793-799.

  [11] YE Y, RICHARD D, LU Y, et al. Measuring daily accessed street greenery: a human-scale approach for informing better urban planning practices[J]. Landscape and Urban Planning, DOI: 10.1016/j.landurbplan.2018.08.028.

  [12] 叶宇, 张灵珠, 颜文涛, 等. 街道绿化品质的人本视角测度框架——基于百度街景数据和机器学习的大规模分析[J]. 风景园林, 2018(8): 24-29.

  [13] SHEN Q, ZENG W, YE Y, et al. StreetVizor: visual exploration of human-scale urban forms based on street views[J]. IEEE Transactions on Visualization and Computer Graphics, 2018, 24(1): 1004-1013.

  [14] BADRINARAYANAN V, KENDALL A, CIPOLLA R. SegNet: a deep convolutional encoder-decoder architecture for image segmentation[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2017(12): 2481-2495.

  [15] YAMAGATA Y, MURAKAMI D, YOSHIDA T. Urban carbon mapping with spatial bigdata[J]. Energy Procedia, 2017, 142: 2461-2466.

  [16] THEIL P. A sequence-experience notation for architectural and urban spaces[J]. Town Planning Review,1961(32): 33-52.

  [17] BANERJEE T, SOUTHWORTH M. City sense and city design. writings and projects of Kevin Lynch[J]. Landscape Journal, 1996, 15(2):167-168,

  [18] TRANCIK R. Finding lost space: theories of urban design[M]. The USA: John Wiley & Sons, 1986.

  [19] BATTY M. Building a science of cities[J]. Cities, 2012, 29(2): S9-S16.

  [20] KATZ P, SCULLY V, BRESSI T W. The new urbanism: toward an architecture of community[M]. New York: McGraw-Hill, 1994.

  [21] MONTGOMERY J. Making a city: Urbanity, vitality and urban design[J]. Journal of urban design, 1998, 3(1): 93-116.

  [22] CHEN L C, PAPANDREOU G, KOKKINOS I, et al. Deeplab: semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected crfs[J]. IEEE transactions on pattern analysis and machine intelligence, 2018, 40(4): 834-848.

  [23] REDMON J, FARHADI A. YOLO9000: better, faster, stronger[C] // 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, 2017: 6517-6525.

  [24] EWING R, HANDY S. Measuring the unmeasurable: urban design qualities related to walkability[J]. Journal of Urban Design, 2009, 14(1): 65-84.

  [25] 叶宇, 庄宇, 张灵珠, 等. 城市设计中活力营造的形态学探究——基于城市空间形态特征量化分析与居民活动检验[J]. 国际城市规划, 2016(1): 26-33.

  [26] YE Y, LI D, LIU X. How block density and typology affect urban vitality: an exploratory analysis in Shenzhen, China[J]. Urban Geography, 2018, 39(4): 631-652.

  [27] YE Y, ZENG W, SHEN Q, et al. The visual quality of streets: a humancentred continuous measurement based on machine learning algorithms and street view images[J]. Environment and Planning B: Urban Analytics and City Science, forthcoming.

  [28] 叶宇, 戴晓玲. 新技术与新数据条件下的空间感知与设计运用可能[J]. 时代建筑, 2017(5): 6-13.

  
  

• 收起