I am developing a sampling technique to do quality assurance checks. One of the things I need for the technique is to find the amount of dead end roads in a certain area. How can this be achieved? I have looked at using JOSM and Overpass-Turbo but I haven't been successful so far. Any help would be appreciated. asked 14 Jul '15, 10:45 BmanS |
You could load the road network data into QGIS and then find dangles answered 14 Jul '15, 15:18 joost schouppe Would also be possible, in QGIS, to find roads that intersect but have no node meaning there is no junction between the roads? and to find all the unnamed roads?
(14 Jul '15, 20:53)
BmanS
Unnamed roads would be straightforward: you could open the attribute table and select roads with an empty name column. Or you can do the same with a simple query. It would require converting your osm data to a format with a limited number of attributes (like for example a shapefile). I'm not sure on the details of finding the missing nodes, but it will probably not be that hard with the "intersect" tool. If you get that far, ask at e.g. gis.stackexchange.com
(15 Jul '15, 07:13)
joost schouppe
|
When you load data into GraphHopper you can use the following Java snippet to calculate all coordinates of dead ends in the whole area and do further analysis:
answered 23 Jul '15, 10:21 peatar |
The pgsnapshot schema from osmosis should be able to help here. You get a SQL version of the OSM data model, and can write SQL queries to find nodes that are only at the start or end of an OSM way with the tags that you want. answered 14 Jul '15, 11:15 rorym |