[Imports] Akros Global Health, offset?

[Nate Smith]

Hi Bjoern - Based on my experience working on that import with Akros, I
would say that matching the DG imagery is best. Bing and DG imagery was
used for the original creation of the data before it was imported. The
offset varied quite a bit across the region and DG was much better.

Nate


Nate Smith
@nas_smith <https://twitter.com/nas_smith>

On May 17, 2017 at 2:43:19 AM, Bjoern Hassler (bjohas+mw at gmail.com) wrote:

Dear all,

I'm looking at this buildings import,
https://wiki.openstreetmap.org/wiki/Northern_Zambia_Buildings_Import

For the buildings It includes buildings near Fiwila, -13.9710, 29.6019,
Zambia.

As well as East (e.g. -13.9762965, 29.5746642) and West (e.g. -13.970185,
29.6203484) of Fiwila.

In Fiwila, the import matches Bing/DG imagery, and the road matches those
too. Hwoever, East and West, there is a clear offset of about 20m East-West.

The wiki page doesn't say how the data was collected. The wiki page states
"The main goal of this work is to not only do a large import of buildings
(268k records) into Zambia, but set an example for how Akros and other
organizations can *use field based operations* to contribute back to
OpenStreetMap. *Data was verified by Akros Global Health contributors with
assistance from the USAID GeoCenter's Remote Sensing Program*." (my bold).

So any thoughts on whether I should go by the Akros data (and offset
Bing/DG) or whether I should adjust the Akros import (in that area) to
match bing?

Many thanks!
Bjoern


_______________________________________________
Imports mailing list
Imports at openstreetmap.org
https://lists.openstreetmap.org/listinfo/imports
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/imports/attachments/20170517/e442ec50/attachment.html>