[GRASS-user] i.cluster and i.maxlik with grayscale image

[Micha Silver]

Markus Neteler wrote:
> On Thu, Jan 1, 2009 at 4:06 PM, Micha Silver <micha at arava.co.il> wrote:
>   
>> Happy New Year to all!
>>     
>
> Happy New Year!
>
>   
>> I want to run some r.le analyses on rasters I will create from two
>>     
>
> [please consider r.li which way faster]
>   
Thanks for the tip.
>   
>> ortho-photo images. The newer ortho-photo is full color, so when importing
>> into GRASS I get the three RGB bands. So I make an image group and subgroup
>> with the three bands and I can do the clustering.  However The older image
>> is a B/W tiff, so it becomes a single grayscale raster.  But the
>> classification modules require a group with more than one band. What's the
>> best way to handle this?
>>     
>
> Note that the i.smap module also works with single-image groups.
>
>   
OK, I'll have a look.
> What means "The older image" above? It's not entirely clear to me
> what you plan to do.
>   
One set of aerial photos are from the 1950's -1960's. These are all 
black and white photos: single band tiff.  The newer set of aerial 
photos are color images from 2005-2007.
The region is a "savanna" area with sparse trees and bushes. We need to 
compare how many trees, and their size, 50 years ago with the situation 
today.

Any additional pointers are most welcome. Thanks,
Micha
> Markus
>
>
>
> This mail was received via Mail-SeCure System.
>
>
>