Hi coworkers
I'm working with grass62. Now I'm trying to develop a code to classify
MODIS images. I have been using i.cluster, but this code perform an
unsupervised classification (in conjunction with i.maxlik). I have three
categories built with v.digit, then I need the signature file for this
training pixels like input to i.maxlik. Someone have developed an
"i.signature" code, with a categorized map like input, and the signature
file like output?
Thanks to all.
Juan Diego
Juan Diego Giraldo Osorio wrote on 09/10/2007 05:17 PM:
Hi coworkers
I'm working with grass62. Now I'm trying to develop a code to classify
MODIS images. I have been using i.cluster, but this code perform an
unsupervised classification (in conjunction with i.maxlik). I have three
categories built with v.digit, then I need the signature file for this
training pixels like input to i.maxlik. Someone have developed an
"i.signature" code, with a categorized map like input, and the signature
file like output?
It is there: i.gensig. Here the overview:
Classification methods in GRASS (table best seen with courier font):
|radiometric,| radiometric, supervised | radio-
and geo-
|unsupervised| |metric,
supervised
--------------+------------+---------------------------------+------------------
Preprocessing | i.cluster |i.class (monitor) i.gensig
(maps)|i.gensigset (maps)
--------------+------------+---------------------------------+------------------
Computation | i.maxlik | i.maxlik i.maxlik | i.smap
Hope this helps,
Markus
------------------
ITC -> dall'1 marzo 2007 Fondazione Bruno Kessler
ITC -> since 1 March 2007 Fondazione Bruno Kessler
------------------
Markus Neteler wrote on 09/10/2007 06:18 PM:
Juan Diego Giraldo Osorio wrote on 09/10/2007 05:17 PM:
Hi coworkers
I'm working with grass62. Now I'm trying to develop a code to classify
MODIS images. I have been using i.cluster, but this code perform an
unsupervised classification (in conjunction with i.maxlik). I have three
categories built with v.digit, then I need the signature file for this
training pixels like input to i.maxlik. Someone have developed an
"i.signature" code, with a categorized map like input, and the signature
file like output?
It is there: i.gensig. Here the overview:
Classification methods in GRASS (table best seen with courier font):
OK, the table didn't survive the formatting.
Just have a look here:
g.manual imageryintro
or
http://grass.itc.it/grass63/manuals/html63_user/imageryintro.html
-> Image classification
Markus
------------------
ITC -> dall'1 marzo 2007 Fondazione Bruno Kessler
ITC -> since 1 March 2007 Fondazione Bruno Kessler
------------------