Life between buildings from a street view image: What do big data analytics reveal about neighbourhood organisational vitality?

Abstract

This article uses big data from images captured by Google Street View (GSV) to analyse the extent to which the built environment impacts the survival rate of neighbourhood-based social organisations in Amsterdam, the Netherlands. These organisations are important building blocks for social life in urban neighbourhoods. Examining these organisations’ relationships with their environment has been a useful way to study their vitality. To extract data on built environment features from GSV images, we applied a deep learning model, DeepLabv3+. We then used elastic net regression to test the relationship between the built environment empirically – distinguishing between car-related, walking-related and mixed-use land infrastructure – and the survival of neighbourhood organisations. This testing approach is novel, to our knowledge not yet having been applied in Urban Studies. Besides revealing the effects of built environment features on the social life between buildings, our study points to the value of easily applicable observational big data. Data captured by GSV and other recently developed methods offer researchers the opportunity to conduct detailed yet relatively swift and inexpensive studies without resorting to overly coarse or common subjective measurements.

Measuring the Built Environment with Google Street View and Machine Learning: Consequences for Crime on Street Segments

“…Although several segmentation models exist in the literature, including FCN8s (Long, Shelhamer, and Darrell 2015), SegNet (Badrinarayanan, Kendall, andCipolla 2017), andPSPNet (Zhao et al 2017), we employed the Deeplabv3+ model (Chen et al, 2018), as it demonstrated excellent accuracy at the Pascal Visual Object Classification challenge as well as on the Cityscapes test dataset, and consistently yielded some of the highest accuracy in various comparisons. For example, Deeplabv3+ exhibited high accuracy (82.1% mIoU) among segmentation models for the Cityscapes test dataset 1 , and 89.0% accuracy on the "PASCAL VOC 2012 datasets", and has therefore been used in several studies for image processing (e.g., Wang and Vermeulen 2020;Liu et al 2019;Du, Ning, and Yan 2020).…”

Built environment and early infection of COVID-19 in urban districts: A case study of Huangzhou

“…The anonymized data about the residential location of infected cases at the initial stage were used to mapping the COVID-19 contagion distribution through geographic information systems (GIS). Various aspects of the urban built environment are correlated with not only organizational level outcomes ( Wang & Vermeulen, 2020 ) but also individual level outcomes regarding population physical and mental health ( Saarloos, Kim, & Timmermans, 2009 ). Using data of the built environment around confirmed cases, we investigated the relationships between the quantity of assembled COVID-19 confirmed cases and the built environment attributes of corresponding neighborhood-level urban space.…”

Big Data in the city

“…Another illustration is provided by Wang and Vermeulen (2021), who explore the role of the built environment in maintaining neighbourhood vitality. Noting that urban design can both enhance and obstruct the potential for collective action, they use machine learning and computer vision algorithms to extract built environment features from images captured by Google Street View (GSV).…”