1

For a city of Radebeul, located in the federal state Saxony in Germany, I am looking for a list with all addresses, something like

id | Street                | Nr  | PLZ   | City
1  | Meißner Straße        | 266 | 01445 | Radebeul
2  | Heinrich-Zille-Straße | 25  | 01445 | Radebeul
.  | ...                   | ... | 01445 | Radebeul

What are those options/sources of data?

I somehow think about deploying overpass-turbo or Nominatim as was suggested in this thread Return all elements associated with address - Overpass API. But perhaps there are other free/open solutions which will somehow differ from OSM data basic.

There was also an idea to extract all points (keeping in mind a certain category/amenity) and all buildings as polygons from OpenStreetMap (e.g. using QGIS, as described in this thread Searching and Downloading OpenStreetMap Data), making geocentoids out of buildings, then defining ($x, $y) for each building feature and each point, and finally put all that coordinates into one common list and proceed via geocoding engine, like HERE Geocoding API or Python Geocoder by means of reverse geocoding.

I am aware of this question is partially related to the "geocoding" topic but I am wondering if I can somehow extract the full list of all addresses, something like a reference list from the geocoding library/solution. Unluckily some resources for Germany were demolished, found in this topic Mass Geocoding Requests for Germany?.

Unfortunately, the Radebeul city portal does not provide such data. Additionally, the data from Deutsche Post probably costs much money.

  • 1
    Nice question. Not sure if the data would come from anyone but the municipality (zB Stadt Zürich), so you may have to ask (bug) them. – philshem May 21 at 10:08
  • Thank you for a hint. I will ask the Stadtverwaltungsamt. Let's see what they gonna reply. Data fro Zürich looks wonderful – Taras May 21 at 10:28
  • 1
    Something really not optimal but doable is to use Google Maps API varying LAT and LONG, if you define a reasonable area and a small LAT LONG variation you will get all addresses. However, it will take time and you'll have to pay for the exceeding requests. – Gustavo Morais May 21 at 12:01
  • @GustavoMorais, if I got you correct you suggest deploying Google Maps Geocoding API, am I right? Regarding the varying tuple {lat, long} what should be the step in your opinion? – Taras May 22 at 5:26
1

It's mentioned in the comments, but in addition to google maps geocoding API, you can use Nominatim geocoding from OSM.

Here's an example query based on latitude and longitude:

https://nominatim.openstreetmap.org/search.php?q=51.108236%2C+13.631143

enter image description here

which resolves to this address:

https://nominatim.openstreetmap.org/details.php?place_id=172430942

Address Tags    
    Radebeul (city)
    266 (housenumber)
    01445 (postcode)
    Meißner Straße (street)

Nominatim has an official API, and some unofficial ones, so in theory you could discretize lat/long pairs on a grid, and the generate N queries, and store results.

Or, you can install your own instance of Nominatim. And then figure out how to extract all addresses within a Polygon from the bulk data. (I'm sure it's a solved problem.)

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.