A methodological investigation of the spatial coverage and temporal variability of Google Street View (GSV) images in small- and medium-sized cities: A people-based approach
Topics:
Keywords: built environment, commute, Google Street View, GeoAI, small- and medium-sized cities, walk
Abstract Type: Paper Abstract
Authors:
Junghwan Kim, Virginia Tech
Kee Moon Jang, Massachusetts Institute of Technology
,
,
,
,
,
,
,
,
Abstract
We conduct a methodological investigation of the spatial coverage and temporal variability of Google Street View (GSV) images adopting a people-based approach. Specifically, we focus on walk commute trajectories (i.e., home-to-work trips) of 97,505 people from 45 small- and medium-sized cities in the U.S., which have been rarely discussed in previous studies. Our results reveal that 44% of commute routes do not have adequate GSV image spatial coverage. Results also demonstrate the substantial variability in their temporal ranges. We reveal that the average monthly variation in timestamps of GSV images on commute trajectories is approximately 7 years, where only about 10% of samples contain GSV images taken within a 1-year period. Lastly, we illustrate regional differences in the spatial coverage scores and temporal variability levels for GSV images of the 45 cities. As our results clearly demonstrate GSV images are imperfect in their spatial coverage and temporal variability, we recommend researchers be aware of these methodological limitations and potential negative impacts on their conclusions.
A methodological investigation of the spatial coverage and temporal variability of Google Street View (GSV) images in small- and medium-sized cities: A people-based approach
Category
Paper Abstract