Peter
I would suggest rectifying before patching, since the two scenes may
have different distortions which should be removed first so that
the overlap will co-register. Rectifying is straightforward if you
use i.rectify. You import the data as an x-y set initially, then
run i.group to gather the different bands together so that they
are rectified as a group, run i.target to select the target (utm)
location, then run i.points to select the registration
points, then i.rectify to move the data from the x-y location to the
utm location. After rectification, use r.patch for each of the bands.
I've never done this for imagery, only rasters, which were imported as
a single band image before rectification, but I think this sequence is
correct.
Good luck
Simon Cox
----
__________________________________________________________________
Dr Simon Cox
__ L
,~' L_|\ Department of Earth Sciences
,-' \ Monash University
( \ Clayton Vic 3168 Australia
\ ___ /
L,~' "\_x/ Phone +61 3 565 5762
u Fax +61 3 565 5062
simon@cerberus.earth.monash.edu.au
__________________________________________________________________
----- End Included Message -----
Dear Sir:
I agree the procedure you have suggested. I have rectified and patched five Landsat MSS images for southeast Alabama. Each Landsat MSS image has four bands. I believe the rectification and patch will be easier for Landsat TM images. It's very difficult to rectify the Landsat MSS images. Anyway, It's over now.
John Wu