Hi, superb job. Congratulations.
I am a geotools/geoserver developer and one of my jobs is fetching
image tiles or the corresponding pyramid tiles from jdbc data
sources. Some weeks ago, I added support for Oracle Georaster and I
want to support WKT Georaster too.
wktraster makes it quite simple to achieve this. Is there any database
layout you recommend. (A table for the image and one for each pyramid,
or put all the tiles into one table having an attribute "level",.....).
I have seen "gdal2wktraster.py", is this the recommended way to import
tiles into postgis ?
What is the table layout created/needed by this utility ?
Cheers
Christian
Quoting Pierre Racine <[email protected]>:
Hi everybody,
I'm pleased to announce that you can now do intersections between
rasters and geometries very much like you used to do
geometry/geometry intersections in PostGIS. For this, PostGIS WKT
Raster introduces two new functions to PostGIS:
ST_Intersects(geometry, raster, band) and ST_Intersection(geometry,
raster, band).
As its geometry/geometry sister, ST_Intersects(geometry, raster,
band) returns TRUE if the withvalue area of a raster or a raster
tile (nodata value are ignored) intersects a geometry and
ST_Intersection(geometry, raster, band) returns the geometry/value
set of geometries representing the intersection between the geometry
and each polygonized group of pixel sharing a same value from the
raster and its associated value. Example:
SELECT polyid,
(ST_Intersection(the_geom, rast)).geom,
(ST_Intersection(the_geom, rast)).val
FROM my_polygons, srtm_tiled
WHERE ST_Intersects(rast, the_geom)
or, a bit more complex, but much faster and returning the same result:
SELECT polyid, (gv).geom, (gv).val
FROM (SELECT polyid, ST_Intersection(the_geom, rast, 1) AS gv
FROM my_polygons, srtm_tiled
WHERE ST_Intersects(rast, the_geom)) foo
These functions works with any kind of geometry so you can intersect
any road, river, polygons or point layer with your favorite
elevation or land cover raster of any resolution and any size. You
can also load any number of raster in the database with one command
line to constitute a unique table raster coverage of any shape (no
matter if your group of raster do not form a rectangular coverage).
You will find a complete tutorial on how to use those two new
functions here:
http://trac.osgeo.org/postgis/wiki/WKTRasterTutorial01. A reference
is available here:
http://postgis.refractions.net/documentation/manual-svn/RT_reference.html
and chapter 13 of "PostGIS in Action" is also a very good
introduction to WKT Raster.
Compilation and installation info are available in the project home
page: http://trac.osgeo.org/postgis/wiki/WKTRaster
These features are the result of two years of collaboration between
many people and companies interested in bringing raster into
PostGIS. I would like to thanks particularly Steve Cumming who
initially made all this possible, Sandro Santilli who wrote the base
code, Mateusz Loskot who wrote the Python loader, Jorge Arevalo who
wrote the GDAL driver and ST_DumpAsPolygons(), Regina and Leo Obe
who are doing so many things and believed in the project from the
beginning by adding a chapter about WKT Raster in "PostGIS in
Action" and also David Zwarg who wrote most of the setter functions.
This is the first step toward a first complete raster/vector
analysis and manipulation SQL API. We hope to make of PostGIS with
WKT Raster the most powerful and complete GIS analysis and
manipulation high level language ever (!!!). WKT Raster should
normally be totally integrated into PostGIS in PostGIS release 2.0.
Upcoming functions should include ST_Reclass, ST_Clip, ST_AsRaster,
ST_Resample.
Let us know your need and your experience with WKT Raster.
Thanks,
Pierre Racine
_______________________________________________
gdal-dev mailing list
[email protected]
http://lists.osgeo.org/mailman/listinfo/gdal-dev
----------------------------------------------------------------
This message was sent using IMP, the Internet Messaging Program.
_______________________________________________
gdal-dev mailing list
[email protected]
http://lists.osgeo.org/mailman/listinfo/gdal-dev
