I am working on crash analysis using OSM data, to analyze predestrain crashes with respect to type of intersections such as:
4-way: X-injuctions
3-way: T-junctions, Y-junctions
multi-leg junctions
Roundabouts
Curves
Long block
I am trying to read: location of intersection, type of road (major/minor) and type of intersection. I noticed it is not straightforward to detect the above intersections because they come in various patterns such as trumpet merges and cloverleafs.
Would like to inquire if there are any existing functions that focus specifically on detecting or classifying such intersection types in urban networks.
Are there known spatial or graph-based techniques (e.g., bearing patterns, topological analysis) that have been used in this context ?
Workaround: After going through the documentation, for the workaround I used the consolidate_intersections() function to simplify the city graph with tolerance=30.
city: Delhi, India
graph type: walk
tolerance: 30
This resulted in nodes at locations where the intersections of any type possibly existed. Though I was not able to get the types of intersections, I got the location and type of roads involved. Further, based on the type of roads, I was able to divide the intersections in the study into 2 types:
major-major & major-minor,
minor-minor
Any suggestions, references, or guidance would be greatly appreciated.