I’m trying to map publicly accessible space based on OSM data. I’m implementing the data analysis in Python. All elements that are of interest to my analysis have to be polygonized at some point, for my visualization. That is particularly challenging with roads, so with way elements with a highway tag. My idea is to assume specific road widths depending on given tags like width, lanes, highway (type), etc. and then to buffer the ways based on these assumptions. I am aware, that this can only be an approximation. I was wondering, if someone developed an approach for polgonizing roads in a previous, similar project.
I want to apply my data analysis and visualization to Vienna, Austria, so I’m also looking for typical road widths in Austria.
Regarding width estimation, maybe you can get some ideas from StreetComplete/RoadWidth.kt at master · streetcomplete/StreetComplete · GitHub
A/B Street - A/B Street also put a LOT of effort into something like this
JOSM/Plugins/measurement - OpenStreetMap Wiki may be useful to tweak parameters
Also, with limited area you may map missing lane/width data, especially in cases where heuristic fail in especially ugly ways
Thank you, that’s all very helpful inspiration!