Utilizing Semantic Segmentation to Analyse Google Street View Imagery for Health-Oriented Urban Planning

After the onset of the Coronavirus Disease 2019 (COVID-19) pandemic, the public health benefits of urban green space (UGS) have attracted increased attention. In exposure science, visibility has been considered a fundamental category for assessing green space exposure, yet evaluations in Denmark rarely address it. This study uses Copenhagen Municipality as a case study to explore the Green View Index (GVI) by analyzing Google Street View (GSV) images using semantic segmentation. The main findings include: (1) a GVI mean of 15.05% within Copenhagen Municipality, ranking below the average of six other European cities, and (2) a map showing the spatial distribution of GVI in the study area. Additionally, this paper considers the potential and possibilities of using street view imagery in future urban health studies.