[OpenLayers-Users] reprojection issue

[Michael O'Cleirigh]

Hi Nick,

I have my layers in a PostGIS database with a structure like:

Original Table:

id integer
zone integer
the_geom geometry

I originally created the table using the 'shp2pgsql' to convert from a 
shapefile.

The process to change the geometry I used was to create a new table with 
the geometry column transformed using the st_transform function:

select id, zone, st_transform (the_geom, 4269) as the_geom into 
degree_projection_table from meter_projection_table_26717 ;

Then I create a new layer in Geoserver from this modified table and 
check what the layer looks like in OpenLayers.

I haven't used either of the tools you mention but its my understanding 
that the st_transform function is standards based so should be available 
via an API.

Regards,

Mike


> Hi Mike,
>
> if you could just tell me which tools to use to do that (GDAL or 
> Shpproj) I'd be really grateful.
> best,
> Nick
>
> On Wed, Jun 10, 2009 at 1:04 PM, Michael O'Cleirigh 
> <michael.ocleirigh at rivulet.ca <mailto:michael.ocleirigh at rivulet.ca>> 
> wrote:
>
>     Hello,
>
>     I had a similar type of problem with WMS  in OpenLayers when
>     showing a layer served through Geoserver.  i.e. The features were
>     offset incorrectly when reprojected onto a Google Maps base layer.
>
>     My source layer is in the NAD 83 UTM Zone 17N EPSG:26917
>     projection. But in order to get it to reproject to the Google
>     EPSG:900913 properly I needed to first convert the layer into
>     standard NAD 83 degree unit projection (4269).
>
>     I see that EPSG:32632 is similar to 26917 in that its unit is
>     meter.  Can you try converting the layer into the appropriate
>     WGS84 degree projection first and see if that helps?
>
>     Regards,
>
>     Mike
>
>>     Hi Jan,
>>
>>     Unless I am very much mistaken,
>>     http://www.dmap.co.uk/utmworld.htm shows Tunisia to be squarely
>>     in UTM 32 N. I have ArcGIS and my landsatt data for the region
>>     all fall into UTM 32 N, which is why I believe this is in fact
>>     the case.
>>
>>
>>     UTM Zone 32 N is also EPSG:32632
>>
>>     http://spatialreference.org/ref/epsg/32632/ so I still don't know
>>     what going on.
>>
>>     UTM coordinates are arbitrary cartesian points, which basically
>>     could end up anywhere which is why the UTM zone is needed for the
>>     origin. The only thing I can think is that yes, there is some
>>     issue with the projection, but I don't know what.
>>
>>     On Wed, Jun 10, 2009 at 12:38 PM, Jan Martin
>>     <janmartin3 at googlemail.com <mailto:janmartin3 at googlemail.com>> wrote:
>>
>>         Hi Nicholas,
>>
>>         your assumption regarding your source UTM Zone beeing UTM 32N
>>         is wrong.
>>
>>         Your coordinates are from Tunisia. And Tunisia is UTM 32S,
>>         not UTM 32N:
>>         http://upload.wikimedia.org/wikipedia/commons/8/85/LA2-Africa-UTM-zones.png
>>
>>         But as you can see here EPSG:32632 is the right result of a
>>         conversion from UTM Zone 32S to EPSG:
>>         http://spatialreference.org/ref/epsg/?search=32632&srtext=Search
>>         <http://spatialreference.org/ref/epsg/?search=32632&srtext=Search>
>>
>>         Question remains why your coordinates from Tunisia are
>>         displayed in Algeria.
>>
>>         I wonder if it is OK to just use EPSG:32632 as a projection
>>         for your data, or if one would have to do math on the
>>         coordinates to.
>>         After all your coordinates have been recorded in *WGS84, UTM
>>         32S.*
>>
>>         Jan
>>
>>         On Wed, Jun 10, 2009 at 6:08 PM, Nicholas Efremov-Kendall
>>         <n.e.kendall at gmail.com <mailto:n.e.kendall at gmail.com>> wrote:
>>
>>
>>             Hello all,
>>
>>             I believe this is more of a gis question, than for
>>             OpenLayers, but the problem manifests itself in
>>             openlayers and I'm having some trouble diagnosing it. Jan
>>             provided me with some very helpful assistance, which led
>>             me to the conclusion that my projection was wrong. I'm
>>             trying to display gps point originally recorded in UTM
>>             32N. I had mistakenly believed this was EPSG 22332, but
>>             36232 is actually what I was going for. My problem is
>>             this: I have the points overlaid on a wms server (thanks
>>             to Jan) but they're in the wrong spot. However, when I
>>             change the EPSG code, the points don't move. Does anyone
>>             have a suggestion for why they don't move? and or for how
>>             to crack this nut.
>>             best,
>>             Nick
>>
>>
>>             _______________________________________________
>>             Users mailing list
>>             Users at openlayers.org <mailto:Users at openlayers.org>
>>             http://openlayers.org/mailman/listinfo/users
>>
>>
>>
>>     ------------------------------------------------------------------------
>>     _______________________________________________ Users mailing
>>     list Users at openlayers.org <mailto:Users at openlayers.org>
>>     http://openlayers.org/mailman/listinfo/users
>
>
>     _______________________________________________
>     Users mailing list
>     Users at openlayers.org <mailto:Users at openlayers.org>
>     http://openlayers.org/mailman/listinfo/users
>
>