Deep Learning from Street View Data

Exposure to natural outdoor environments including green space is vital for people’s health. Remote sensing via satellites is most used to generate measures of green space, however this does not capture the street-level perspective that people experience.

We addressed this research challenge with an assemblage of IT methods to derive measures of greenery from street view services on the web which allows virtual navigation through urban spaces composed of geo-tagged street-level images. We developed methods to automatically compute objective and accurate measures of green space employing big street view data and deep learning. 

Researcher

Marco Helbich, associate professor, Human Geography and Spatial Planning 

Associated researchers

Daniel Oberski (Methodology and Statistics, Faculty of Social and Behavioural Sciences), Ronald Poppe (Information and Computing Sciences, Faculty of Science), Maarten Zeylmans van Emmichoven (Physical Geography, Faculty of Geosciences), Raoul Schram (Data Engineering, Information and Technology Services)

Partners

ERC, UU IT Innovation Fund