[OSM-dev] indoor wifi geopositionning - openstreetmap precision, collaborators?
richard at 2009.atterer.net
Wed Dec 16 21:39:13 GMT 2009
thanks for CCing me, Jonathan-David! It's great to see other people are
interested in indoor navigation! I just subscribed to the dev list.
On Wed, Dec 16, 2009 at 10:50:01AM +0100, Jonathan-David SCHRODER wrote:
> Outside of openstreetmap's site, we have found several people mentioning the
> topic of indoor geopositioning :
> - Richard Atterer here : http://atterer.net/leadme ;
I just extended that page with pictures of the hardware I built for indoor
mapping, and a screenshot of the map editor I'm developing.
(I should mention right now that my time to continue working on this is
very constrained in the nearer future. :~-| )
> - on mobile phones : use of Nutiteq's j2me maps library, or other
> android piece of open source software for example (osmdroid, opensatnav).
> Some code around those apps allows for more zoom levels & for
> threaded-background gathering of wifi beacons positions (stored in 2)'s
> database) so as to make triangulation calculation and to allow end-user
> searching of objects to display on top of map (what do you propose for this
> overlaying ? Can the Web API reply to the queries with a KML group of
> features ?... we want to have features's URLs known and usable on the mobile
>From what I gathered (from researchers working in this area), triangulation
in an indoor setting is problematic. Signals may be attenuated due to
walls, or may be reflected from walls, so it just does not work well.
To get around this issue, my simple idea was to continuously record WiFi
beacons and their relative strength, and include many many samples in the
map - with every node, and possibly even in extra nodes created just to
carry that information. With a decent algorithm, this might give better
results than storing AP location info in the map.
| ) /| Richard Atterer
| \/ | http://atterer.net
More information about the dev