[Tagging] [OSM-talk-be] cadastral plan now open data

André Pirard A.Pirard.Papou at gmail.com
Sat Sep 22 00:08:26 UTC 2018

On 2018-09-22 01:08, Lionel Giard wrote:
> André, i don't really see your point with the argument you made about 
> cadastre data. i don't know of anybody looking to use this data to add 
> building into OSM. It is outdated data regarding buildings and roads 
> in a lot of different place in Belgium. And as, the people at cadastre 
> are not the source for these data, they don't need to update it 
> anymore !  Most people are only using the default imagery (they often 
> don't know about anything else).
> When tracing buildings, we should always base our self on the PICC for 
> Wallonia or Urbis for Brussels or GRB for Flanders and the relevant 
> best imagery in each case (comparing them keeping in mind that 
> vectorial data are not up to date everywhere at the same time).

I'm not sure what you read because what you say is exactly what I wrote:
> But there is a *BIG WARNING*.*
> *PLEASE *DO NOT* use the [Cadastre] WMS to trace, and I suppose that 
> the import data exhibits the same problem.
> ..
> For what I have seen of it in Wallonia, the Cadastre WMS map usually 
> exhibits an error shift (wrong coordinates) as large as 10m and the 
> houses and other elements have a wrong shape and location within the 
> parcel.
> See example below.
> Instead, use JOSM with PICC in Wallonia and AGIV in Flanders.
> The Cadastral map can be useful to:
>   * find a house number not displayed on other maps
>   * find a house not displayed on other maps, but do not rely on the
>     house location (latlon)
>   * trace boundaries, especially former municipalities', but be aware
>     that Cadastre's boundaries are slightly different from other
>     sources, in particular because they avoid crossing a parcel. They
>     are official.
>   * use names not displayed on other maps and in particular a very
>     nice set of locality names
>   * maybe use other features not needing a precise location
> So, the usage of the cadastre is limited to displaying a JOSM layer, 
> just for comparison with the PICC, aerial or other maps.
> Even more than tracing, beware of importing cadastral data !
> PLEASE DO NOT destroy with cadastre imports patient corrections with 
> JOSM and PICC of ID and Potlatch inaccuracies and mistakes !!!
What's the problem?
I'm glad that we totally agree.

About what you write below, if you have contacts with PICC (SPW don't 
always reply)...
You don't define ICAR (not free) but I seem to recall that there is a 
free list of street names.
Could you ask them to specify the coordinates of both ends in this list?
I think that the list is freely usable and those coordinates would allow 
OSM to at last start mapping roads correctly.

Also, PICC welcome corrective remarks but they don't say how to do that 
and they ... don't reply.
I have quite a number of corrections I didn't send because of that.
I suggested reviving OpenStreetBugs or equivalent for OSM to notify bugs 
to PICC ... or equivalent ... no answer.
Support in JOSM would be nice.

All the best,


> To quote a report from the last PICC meeting in June, wallonia have 
> the following idea on official source :
>     Wallonia is planning a “*georeferentiel*” (base of layers that are
>     high accuracy and authentic source) within the application of
>     INSPIRE.There are 4 major datasets that will be made as accurate
>     as possible and with one authentic source for each:
>       * /buildings/ (in PICC),
>       * /admin limits/ (in cadastre),
>       * /addresses/ (in ICAR) and
>       * /Roads/ (in PICC/”direction des routes” collaboration with
>         attributes (like primary, secondary, highway, bridge,…) on
>         road, not only geometry).
> from: 
> https://wiki.openstreetmap.org/wiki/WikiProject_Belgium/Contacts_with_local_autorities/Wallonia/Compte_rendu_Rencontre_des_utilisateurs_du_PICC_Juin_2018
> So cadastre is really about admin. limits and parcels :
> - For admin. limits i already talked about it before (as quoted, it is 
> even the authentic source for all Belgium apparently),
> - and for parcels, there is an open debate about adding this 
> information to OSM (mainly because it is huge amount of work to 
> maintain it) -> as discussed on the wiki page 
> https://wiki.openstreetmap.org/wiki/Parcel
> So there is no ambiguity in what official data exist, the only problem 
> is that ICAR for the addresses is not yet finished for all 
> municipalities (and not complete in the PICC as far as i know ?! but i 
> may be wrong about that) and more importantly, it is not open nor 
> viewable except for special customers. So we still need to rely on 
> tracing from the imagery and the PICC WMS at the moment (for 
> buildings, roads and addresses) and now, the Cadastre data (or wms) to 
> maybe update admin limits.
> Note that these official data are considerably improving over time ! 
> They are still outdated or imprecise in some place, but even then it 
> is still the official source at the moment (so we should NOT copy 
> blindly as you said, but compare with recent imagery and survey !!!). 
> The plan is to make an update cycle closer to 1-2 years (for every 
> municipalities) for the PICC, but they are still far from it (6.5 
> years of average update time).
> Is there something i'm missing out in your explication ? If so tell 
> me, i may not understand all your point !
> Le ven. 21 sept. 2018 à 23:23, Martin Koppenhoefer 
> <dieterdreist at gmail.com <mailto:dieterdreist at gmail.com>> a écrit :
>     sent from a phone
>     On 21. Sep 2018, at 21:32, André Pirard <A.Pirard.Papou at gmail.com
>     <mailto:A.Pirard.Papou at gmail.com>> wrote:
>>     The whole cadastral map is offset by that 7m.
>>     ...
>>     *Picture 3:* So, I dragged his parcel right onto the wall.
>>     And now it's correctly located, aligned with the fencing all around.
>     how did you know which source was off, the cadastral map or the
>     orthophoto?
>>     ...
>>     The Belgian cadastre is not the only one with an error shift.
>>     With JOSM, I have similarly proved that Google Map has a 120m NE
>>     shift in Beijing.
>>     Nobody noticed it.
>     it is well known that the chinese government requires all imagery
>     and map providers to use chinese algorithms which distort the map
>     coordinates systematically, in a way that they remain usable as
>     long as your navigation system uses the same algorithms.
>     Ciao, Martin
>     _______________________________________________
>     Tagging mailing list
>     Tagging at openstreetmap.org <mailto:Tagging at openstreetmap.org>
>     https://lists.openstreetmap.org/listinfo/tagging
> _______________________________________________
> Tagging mailing list
> Tagging at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/tagging

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20180922/60026dff/attachment.html>

More information about the Tagging mailing list