[GRASSLIST:4139] Re: Map import cluelessness

Ian MacMillan ian_macmillan at umail.ucsb.edu
Mon Aug 9 15:12:36 EDT 2004


So there are a couple of points to consider here.  First, I think you 
forget to mention the program i.points (which is the one you are 
interested in).  This program is the one that sets the points used to 
convert the data from a non-georeferenced scanned map, into a 
full-fledged georectified map.  To use it, you first must use i.group 
(sets up your imagery group), then i.target (this says where you want 
your scanned map projected to), then i.points.

This command brings up an interface in your monitor where you select 
points to register.  There are two ways to do this.  First (using 
keyboard), you can click a point on your scanned map, then enter the 
coordinates of this point in the format of the projection you want.  So 
for your case, you are going to enter in something like
25.78 45.80
this is your latitude and longitude in your target location (NOT the xy 
coordinates that are given back to you on screen).  You would typically 
get these points from lat-long that is on your scanned map, for example 
if you have a grid with degrees on your map.

The other method (using screen) is only valid if you have a map that 
overlaps in space with the scanned map in your target location.  You 
can then use this already georeferenced map to georeference your 
scanned map.  Select "Plot Raster" at the bottom of the screen, then 
choose the map you want from the target location.  You can then click a 
point on your map, then click the corresponding point on the map from 
the target location.  This will automatically force the two maps to 
match at the points you select.

After this process is done (for a minimum of 4 points), you can check 
how much distortion will occur in your scanned map by clicking 
"analyze".  The larger the error, the more distorted your image will 
be.  You can delete points in here if they seem wrong, then add new 
ones later.

After this, run i.rectify, aftger your processor stops working, the 
image will be in your target location.  Open up the new location (after 
exiting from your xy location), and you are good to go.

Good luck,
Ian


On Aug 9, 2004, at 11:35 AM, C.S. Cornuelle wrote:

> I am trying to import a scanned map into GRASS 5.3 and transform it 
> into
> the same coordinate system/projection as a grass database, e.g. a
> DEM in lat-lon.
>
> Having read the Spearfish tutorial and the Imagery tutorial, and having
> flailed away at my data, it occurs to me that I remain fairly
> clueless.  As I dimly understand it, one dumps the scanned map into
> a grass xy database.  Then i.group, i.target, and i.rectify are used
> to transform this into a georeferenced database which is in the same
> projection as the paper map, e.g. mercator.  Finally one would use
> r.proj to import this into the "target" database in which one really
> wants to work, e.g. lat-lon.
>
> Is that correct?
>
> What I am not understanding is how to use i.target (or other tools) to
> achieve this.  Initially it seems that one would need to mark with
> the cursor a location, e.g. a city, on the raw xy data, and then
> enter into the software the corresponding coordinates, e.g. latitude 
> and
> longitude, for the scanned map's actual projection, e.g. mercator.  But
> on the line
>
> 	Enter coordinates as east north:
>
> I am unclear on what to enter.  It seems that one should just type in
> the xy appearing on the screen, e.g.
>
> 	Point 2 marked on image at
>
> 	  East:     7342.50
> 	  North:    5578.50
>
> But how can this move things forward?  Should not i.target ask for the
> corresponding georeferenced coordinates, e.g. the latitude and 
> longitude
> of Mt. Olympus Utah USA?  Would not i.rectify need that information 
> from
> outside raw xy map in order to generate a new map, e.g. in the mercator
> projection/coordinate system?
>
> Thanks for your patience, as I'm only a physicist and am easily
> confused.  :^)
>
> -- 
> Adios,
> Chris Cornuelle
> bob at xmission dot com
>




More information about the grass-user mailing list