Deep learning from Street View data
Deep learning from Street View data
Spending time in natural outdoor settings, especially in areas with green spaces, is essential for people’s health. Research indicates that such exposure can have positive effects on mental well-being, stress reduction, and overall physical health.
A common method used in environmental research is remote sensing via satellites. However, while satellites provide a broad overview, they don't capture the detailed, street-level perspective that individuals experience in their daily lives. In other words, relying solely on satellite imagery might miss the more localized and nuanced aspects of green spaces. It doesn't account for the specific ways people interact with and perceive greenery on the ground level, such as in parks, gardens, or along streets.
In this research, we use a combination of Information Technology (IT) methods, such as street view services that allow users to virtually navigate through streets using geo-tagged street-level images, providing a realistic representation of urban spaces. Through algorithms and computational processes, we analyse the street-level images and quantify the extent and characteristics of greenery. These methods enhance the accuracy and efficiency of measuring green spaces.
Researcher
Dr. Marco Helbich
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