Dataset Information

Predicting walking-to-work using street-level imagery and deep learning in seven Canadian cities.

ABSTRACT: New 'big data' streams such as street-level imagery are offering unprecedented possibilities for developing health-relevant data on the urban environment. Urban environmental features derived from street-level imagery have been used to assess pedestrian-friendly neighbourhood design and to predict active commuting, but few such studies have been conducted in Canada. Using 1.15 million Google Street View (GSV) images in seven Canadian cities, we applied image segmentation and object detection computer vision methods to extract data on persons, bicycles, buildings, sidewalks, open sky (without trees or buildings), and vegetation at postal codes. The associations between urban features and walk-to-work rates obtained from the Canadian Census were assessed. We also assessed how GSV-derived urban features perform in predicting walk-to-work rates relative to more widely used walkability measures. Results showed that features derived from street-level images are better able to predict the percent of people walking to work as their primary mode of transportation compared to data derived from traditional walkability metrics. Given the increasing coverage of street-level imagery around the world, there is considerable potential for machine learning and computer vision to help researchers study patterns of active transportation and other health-related behaviours and exposures.

SUBMITTER: Doiron D

PROVIDER: S-EPMC9626470 | biostudies-literature | 2022 Nov

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Predicting walking-to-work using street-level imagery and deep learning in seven Canadian cities.

Doiron Dany D Setton Eleanor M EM Brook Jeffrey R JR Kestens Yan Y McCormack Gavin R GR Winters Meghan M Shooshtari Mahdi M Azami Sajjad S Fuller Daniel D

Scientific reports 20221101 1

New 'big data' streams such as street-level imagery are offering unprecedented possibilities for developing health-relevant data on the urban environment. Urban environmental features derived from street-level imagery have been used to assess pedestrian-friendly neighbourhood design and to predict active commuting, but few such studies have been conducted in Canada. Using 1.15 million Google Street View (GSV) images in seven Canadian cities, we applied image segmentation and object detection com ...[more]

PMID: 36319661

Similar Datasets

Project description:BackgroundStreet imagery is a promising and growing big data source providing current and historical images in more than 100 countries. Studies have reported using this data to audit road infrastructure and other built environment features. Here we explore a novel application, using Google Street View (GSV) to predict travel patterns at the city level.MethodsWe sampled 34 cities in Great Britain. In each city, we accessed 2000 GSV images from 1000 random locations. We selected archived images from time periods overlapping with the 2011 Census and the 2011-2013 Active People Survey (APS). We manually annotated the images into seven categories of road users. We developed regression models with the counts of images of road users as predictors. The outcomes included Census-reported commute shares of four modes (combined walking plus public transport, cycling, motorcycle, and car), as well as APS-reported past-month participation in walking and cycling.ResultsWe found high correlations between GSV counts of cyclists ('GSV-cyclists') and cycle commute mode share (r = 0.92)/past-month cycling (r = 0.90). Likewise, GSV-pedestrians was moderately correlated with past-month walking for transport (r = 0.46), GSV-motorcycles was moderately correlated with commute share of motorcycles (r = 0.44), and GSV-buses was highly correlated with commute share of walking plus public transport (r = 0.81). GSV-car was not correlated with car commute mode share (r = -0.12). However, in multivariable regression models, all outcomes were predicted well, except past-month walking. The prediction performance was measured using cross-validation analyses. GSV-buses and GSV-cyclists are the strongest predictors for most outcomes.ConclusionsGSV images are a promising new big data source to predict urban mobility patterns. Predictive power was the greatest for those modes that varied the most (cycle and bus). With its ability to identify mode of travel and capture street activity often excluded in routinely carried out surveys, GSV has the potential to be complementary to new and traditional data. With half the world's population covered by street imagery, and with up to 10 years historical data available in GSV, further testing across multiple settings is warranted both for cross-sectional and longitudinal assessments.

Project description:BackgroundAlthough walking for travel can help in reaching the daily recommended levels of physical activity, we know relatively little about the correlates of walking for travel in the European context.ObjectiveWithin the framework of the European Physical Activity through Sustainable Transport Approaches (PASTA) project, we aimed to explore the correlates of walking for travel in European cities.MethodsThe same protocol was applied in seven European cities. Using a web-based questionnaire, we collected information on total minutes of walking per week, individual characteristics, mobility behavior, and attitude (N=7,875). Characteristics of the built environment (the home and the work/study addresses) were determined with geographic information system (GIS)-based techniques. We conducted negative binomial regression analyses, including city as a random effect. Factor and principal component analyses were also conducted to define profiles of the different variables of interest.ResultsLiving in high-density residential areas with richness of facilities and density of public transport stations was associated with increased walking for travel, whereas the same characteristics at the work/study area were less strongly associated with the outcome when the residential and work/study environments were entered in the model jointly. A walk-friendly social environment was associated with walking for travel. All three factors describing different opinions about walking (ranging from good to bad) were associated with increased minutes of walking per week, although the importance given to certain criteria to choose a mode of transport provided different results according to the criteria.DiscussionThe present study supports findings from previous research regarding the role of the built environment in the promotion of walking for travel and provides new findings to help in achieving sustainable, healthy, livable, and walkable cities. https://doi.org/10.1289/EHP4603.

Project description:Street food makes a significant contribution to the diet of many dwellers in low- and middle-income countries and its trade is a well-developed activity in the central Asian region. However, data on its purchase and nutritional value is still scarce. This study aimed to describe street food purchasing patterns in central Asia, according to time and place of purchase. A multicentre cross-sectional study was conducted in 2016/2017 in the main urban areas of four central Asian countries: Dushanbe (Tajikistan), Bishkek (Kyrgyzstan), Ashgabat (Turkmenistan) and Almaty (Kazakhstan). Street food markets (n = 34) and vending sites (n = 390) were selected by random and systematic sampling procedures. Data on the purchased foods and beverages were collected by direct observation. Time and geographic location of the purchases was registered, and their nutritional composition was estimated. A total of 714 customers, who bought 852 foods, were observed. Customers' influx, buying rate and purchase of industrial food were higher in city centers compared to the outskirts (median: 4.0 vs. 2.0 customers/10 min, p < 0.001; 5.0 vs. 2.0 food items/10 min, p < 0.001; 36.2 vs. 28.7%, p = 0.004). Tea, coffee, bread and savory pastries were most frequently purchased in the early morning, bread, main dishes and savory pastries during lunchtime, and industrial products in the mid-morning and mid-afternoon periods. Energy and macronutrient density was highest at 11:00-12:00 and lowest at 09:00-10:00. Purchases were smaller but more energy-dense in city centers, and higher in saturated and trans-fat in the peripheries. This work provides an overview of the street food buying habits in these cities, which in turn reflect local food culture. These findings from the main urban areas of four low- and middle-income countries which are currently under nutrition transition can be useful when designing public health interventions customized to the specificities of these food environments and their customers.

Dataset Information

Predicting walking-to-work using street-level imagery and deep learning in seven Canadian cities.

Publications

Predicting walking-to-work using street-level imagery and deep learning in seven Canadian cities.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets