[OSM-dev] Projection problems
grand.edgemaster at gmail.com
Tue Dec 16 17:27:43 GMT 2008
osm2pgsql uses a spherical mercator projection which cuts down on the
amount of math required to produce a reasonable estimation of the
earth for our uses.
It's not a 'proper' projection in the GIS sense, but it works for our purposes.
I'm not sure what scaling factor osm2pgsql uses
Somewhat related implementations are at
2008/12/16 Bradley Kite <bradley.kite at gmail.com>:
> Hi all,
> I wonder if any one could be so kind as to help me out with a problem
> I am having with projections:
> I am trying to convert from lat/long coords to the Mercator projection
> as is used in osm2pgsql
> I am using this code:
> The Lat/Long I am using is (bounding box around a street I know):
> -0.90362548828125,51.38892310050485 to -0.90087890625,51.38720911696496
> However, when converted to Mercator projection as per the above Java
> code, I get the following:
> -100591.12922329194,6656970.686752424 to -100285.38111015123,6656665.738991335
> The y values (6656970 - 6656665) do not match up with data the
> postgres database:
> LINESTRING(-100491.52 6690303.05,-100474.32 6690280.11,-100456.16
> 6690259.1,-100438.02 6690235.21,-100434.02 6690222.73,-100430.37
> 6690211.33,-100433.23 6690176.94,-100434.19 6690146.36,-100432.48
> 6690134,-100430.37 6690118.64,-100417.95 6690093.81,-100398.83
> 6690072.78,-100373.27 6690047.38,-100370.16 6690038.6,-100327.34
> 6689918.14,-100309.98 6689869.27,-100245.01 6689884.55)
> Note - all the Y values in the DB are ~6690000 not ~6650000 as
> returned by the mercY() function.
> The effect of this is that an SQL query to pull data from the database
> is not fetching the correct data.
> Would there happen to be an expert on this list that help out with the
> accuracy of the projection code at the above link? It only affects the
> Y coordinate, X seems fine.
> Many thanks in advance.
> dev mailing list
> dev at openstreetmap.org
More information about the dev