[GRASSLIST:2710] RE: i.rectify and image resolution

Beaudry Charles (NORANDA) BeaudryC at normin.com
Thu Nov 29 09:21:37 EST 2001


Ulf Mehlig,


> 
>  CURRENT REGION: N=1600  S=0  RES=1  ROWS=1600
> 		 E=1600  W=0  RES=1  COLS=1600
>  PROJECTION: 0 (x,y)
>  ZONE:       0

I don't know how you are importing the image but if it is in some non-grid
format the above is OK if your image has 1600 pixcels wide and 1600 pixels
high or less.  If the image is bigger you will get truncation.  



> Now, I want to rectify the image. I created a target location in
> lat/long-coordinate system with the highest resolution I could set,
> and the "sam69" spheroid (my image is from equatorial Brazil -- is
> this the right choice?). 
> 
>  CURRENT REGION: N= 0:43:48S  S=1:12:36S  RES=0:00:00.000036  
> ROWS=48000000
> 		 E=46:55:48E  W=  46:27E  RES=0:00:00.000036  
> COLS=48000000
>  PROJECTION: 3 (Latitude-Longitude)
>  ZONE:       0
>  DATUM:      ** unknown (default WGS84) **
>  ELLIPSOID:  sam69
> 

TM7 images are 30 m in resolution which, in degrees is about 1 sec (RES =
0:00:00.972) or in decimal circa 0.00027 degrees.  The resolution of your
location appears small compared to the resolution of the image.


> Then I made an image group, set the target location, and marked points
> via i.points (I have tried a set of 8-12 GPS positions of points
> +-well recognizable in the image).
> 
> Afterwards, I called i.rectify. I specified "choose smallest region
> that covers the image" but I was not able to set a higher resolution
> than 0.01. The resulting image in the targe location has consequently
> only e very rough raster: 
> 
>  CURRENT REGION: N= 0:43:48S  S=1:12:36S  RES=0:00:36  ROWS=48
> 		 E=46:55:48E  W=  46:27E  RES=0:00:36  COLS=48
>  PROJECTION: 3 (Latitude-Longitude)
>  ZONE:       0
>  DATUM:      ** unknown (default WGS84) **
>  ELLIPSOID:  sam69
> 

I have had problems with the SAM69 datum.  There appears to be a confusion
in the ellipsoid and datum files.  I fixed it but don't have it with me.  I
will get it to Markus tonight for uploading to CVS and send a copy directly
to you.

I would set up the right projection and set the resolution of my destination
location to 0:00:0.972 or 0.00027 in decimal before trying to run i.rectify.

Charles Beaudry
-------------- next part --------------
-----

This message has been scanned by the AT&T Global Network Services Virus Detect Service. Ce message a ?t? scrut? par le service Virus Detect des Services de r?seaux mondial AT&T.

-----


More information about the grass-user mailing list