[postgis-users] Rasters - world wide SRTM dataset for altitude profile?

29 views
Skip to first unread message

Frederik Ramm

unread,
Feb 14, 2012, 6:54:54 AM2/14/12
to PostGIS Users Discussion
Hi,

forgive me for barging in with little background knowledge. I'm a
regular PostGIS user but haven't used rasters with PostGIS yet, and I
was hoping that you could give me a quick "reality check" on the
following idea.

In OpenStreetMap we sometimes have the requirement to generate altitude
profiles for a route. There's a 4-year-old summer of code project that
imports SRTM data into Postgres, by creating a two-column table of
quadtile location code (bigint) and altitude (int), and then
step-by-step assembling height values for a given route geometry.

Importing SRTM data into that structure takes quite long and needs a lot
of storage space.

I wonder if, with PostGIS raster support, I could:

* simply import a world-wide SRTM data set (I'm sure people must have
done that already?)
* and then efficiently query the nearest raster values for a given
linestring geometry, thereby getting an altitude profile?

Without having checked the limits of PostGIS raster support, it sounds
to me as if this could be much simpler than the old approach. Can you
see any obvious problems with that idea, or should I go ahead and try it?

Bye
Frederik

_______________________________________________
postgis-users mailing list
postgi...@postgis.refractions.net
http://postgis.refractions.net/mailman/listinfo/postgis-users

Sandro Santilli

unread,
Feb 14, 2012, 9:00:15 AM2/14/12
to PostGIS Users Discussion
On Tue, Feb 14, 2012 at 12:54:54PM +0100, Frederik Ramm wrote:

> I wonder if, with PostGIS raster support, I could:
>
> * simply import a world-wide SRTM data set (I'm sure people must
> have done that already?)
> * and then efficiently query the nearest raster values for a given
> linestring geometry, thereby getting an altitude profile?

Yes and yes !

> Without having checked the limits of PostGIS raster support, it
> sounds to me as if this could be much simpler than the old approach.
> Can you see any obvious problems with that idea, or should I go
> ahead and try it?

I think you should be fine to go ahead and try it.
Let us know how it goes !

--strk;

Reply all
Reply to author
Forward
0 new messages