[GRASS-user] how to create multiband/hyperspectral

[Michael Barton]

Robert,

It isn't clear what you want to do. If you want to patch together adjacent
images (i.e., rasters covering different geographic areas), use r.patch.

If you want to make a color image out of 3 bands, use d.rgb to just display
a color image, or r.composite to combine 3 bands into a new color image
file.

GRASS stores all individual 'bands' as separate rasters that you can
combine, analyze, and process in multiple ways. It does not combine multiple
bands into a single 'file'.

Some processes can deal with a group of images/bands covering the same
geographic area as a 'group'. A group is created with i.group. But a 'group'
is simply a reference to a set of images; it does not combine them into a
single multi-band file. This is unnecessary with GRASS.

Michael
__________________________________________
Michael Barton, Professor of Anthropology
School of Human Evolution & Social Change
Center for Social Dynamics & Complexity
Arizona State University

phone: 480-965-6213
fax: 480-965-7671
www: http://www.public.asu.edu/~cmbarton



> From: Robert Kuszinger <kuszinger at giscom.hu>
> Date: Tue, 11 Jul 2006 06:59:59 +0200
> To: <grassuser at grass.itc.it>
> Subject: [GRASS-user] how to create multiband/hyperspectral
> 
> 
> Hello!
> 
> I give up - cannot find the solution. :)
> 
> I've created a python script to automatically process ASTER CD-s and
> create all the raster layers in grass from a given CD automatically. Now
> I have a LOT of raster layers.
> 
> *** I would like to join raster layers covering the same area *** to
> make only one file for one area. So, how to make a multispectral file
> using several raster layers? I have 3-14 bands I'd like to join per image
> How to make it?
> 
> I checked grass documentation, commands, but simply cannot figure how to
> make it.
> I realized that there is an i.group command - that could be fine for
> rasters as well...
> 
> Maybe the answer is too simmple, that's why I can't find.
> 
> Thanks for any help
> 
> Robert
> 
> 
> 
> 
> 
> 
> 
>