CaltechAUTHORS
  A Caltech Library Service

Classification and mapping of urban canyon geometry using Google Street View images and deep multitask learning

Hu, Chuan-Bo and Zhang, Fan and Gong, Fang-Ying and Ratti, Carlo and Li, Xin (2020) Classification and mapping of urban canyon geometry using Google Street View images and deep multitask learning. Building and Environment, 167 . Art. No. 106424. ISSN 0360-1323. https://resolver.caltech.edu/CaltechAUTHORS:20191004-105018148

[img] PDF - Published Version
See Usage Policy.

3876Kb

Use this Persistent URL to link to this item: https://resolver.caltech.edu/CaltechAUTHORS:20191004-105018148

Abstract

Urban canyon classification plays an important role in analyzing the impact of urban canyon geometry on urban morphology and microclimates. Existing classification methods using aspect ratios require a large number of field surveys, which are often expensive and laborious. Moreover, it is difficult for these methods to handle the complex geometry of street canyons, which is often required by specific applications. To overcome these difficulties, we develop a street canyon classification approach using publicly available Google Street View (GSV) images. Our method is inspired by the latest advances in deep multitask learning based on densely connected convolutional networks (DenseNets) and tailored for multiple street canyon classification, i.e., H/W-based (Level 1), symmetry-based (Level 2), and complex-geometry-based (Level 3) classifications. We conducted a series of experiments to verify the proposed method. First, taking the Hong Kong area as an example, the method achieved an accuracy of 89.3%, 86.6%, and 86.1%, respectively for the three levels. Even using the field survey data as the ground truth, it gained approximately 80% for different levels. Then, we tested our pretrained model in five other cities and compared the results with traditional methods. The transferability and effectiveness of the scheme were demonstrated. Finally, to enrich the representation of more complicated street geometry, the approach can separately generate thematic maps of street canyons at multiple levels to better facilitate microclimatic studies in high-density built environments. The developed techniques for the classification and mapping of street canyons provide a cost-effective tool for studying the impact of complex and evolving urban canyon geometry on microclimate changes.


Item Type:Article
Related URLs:
URLURL TypeDescription
https://doi.org/10.1016/j.buildenv.2019.106424DOIArticle
ORCID:
AuthorORCID
Zhang, Fan0000-0002-3643-018X
Gong, Fang-Ying0000-0001-5194-6435
Additional Information:© 2019 Elsevier Ltd. Received 17 May 2019, Revised 21 September 2019, Accepted 23 September 2019, Available online 3 October 2019. The funding support for this research is provided by the National Natural Science Foundation of China under Grant 41901321. Chuanbo Hu and Xin Li's work is partially supported by the DoJ/NIJ under grant NIJ 2018-75-CX-0032, NSF under grant OAC-1839909 and the WV Higher Education Policy Commission Grant (HEPC.dsr.18.5). The authors would also like to gratefully thank the members of the MIT Senseable City Lab Consortium: RATP, Dover Corporation, Teck Resources, Lab Campus, Anas S.p.A., Ford, SNCF Gares & Connexions, Brose, Allianz, ENEL Foundation, Laval, Curitiba, Stockholm, Amsterdam, Victoria State Government, KTH Royal Institute of Technology, UTEC - Universidad de Ingenier–a y Tecnología, Politecnico di Torino, Austrian Institute of Technology, Fraunhofer Institute, Kuwait-MIT Center for Natural Resources, SMART - Singapore-MIT Alliance for Research and Technology, and AMS Institute for supporting this research. Declaration of competing interest: The authors declare that there is no conflict of interest.
Funders:
Funding AgencyGrant Number
National Natural Science Foundation of China41901321
Department of Justice2018-75-CX-0032
NSFOAC-1839909
West Virginia Higher Education Policy CommissionHEPC.dsr.18.5
Subject Keywords:Street canyon classification; Built environment; Aspect ratio; Deep learning; Google street view
Record Number:CaltechAUTHORS:20191004-105018148
Persistent URL:https://resolver.caltech.edu/CaltechAUTHORS:20191004-105018148
Official Citation:Chuan-Bo Hu, Fan Zhang, Fang-Ying Gong, Carlo Ratti, Xin Li, Classification and mapping of urban canyon geometry using Google Street View images and deep multitask learning, Building and Environment, 2019, 106424, ISSN 0360-1323, https://doi.org/10.1016/j.buildenv.2019.106424. (http://www.sciencedirect.com/science/article/pii/S0360132319306341)
Usage Policy:No commercial reproduction, distribution, display or performance rights in this work are provided.
ID Code:99086
Collection:CaltechAUTHORS
Deposited By: Tony Diaz
Deposited On:04 Oct 2019 17:58
Last Modified:15 Oct 2019 16:35

Repository Staff Only: item control page