Yes it is quite possible.
You may want to take a look at Lizarazo and Elsner paper in
International Journal of Remote Sensing entitled Fuzzy Segmentation for
Object-based Image Classification. It is my understanding that the
methodology they describe was done almost entirely within R.
On Tue, Aug 10, 2010 at 8:19 AM, Chethan S <[email protected]> wrote:
> Hello everybody!
>
> For my academic project work I intend to use only open-source software. My
> objective is to carry out classification of satellite digital images using
> Geostatistics. I found out GRASS is a good software for image processing.
> Also a good geostatistics software is R. It is also known that R can be used
> in combination with GRASS. But I am not able to confirm that satellite
> digital image classification can be carried out using geostatistical
> analysis with GRASS in combination with R.
>
> Can anybody confirm the capability of the software in question or give any
> alternative suggestions?
Yes, this can be confirmed. We are currently peer-reviewing a related paper
for Springer Lecture Notes on the OGRS 2009 conference in Nantes,
You can find the conference contribution here:
http://www.ogrs2009.org/doku.php?id=research
"An alternative approach for object oriented classification of historical
black & white aerial photographs using GRASS GIS and R"
We are working here in my institution on a similar project and will likely
document the steps in the GRASS Wiki in the near future.
Markus
--
http://gis.ambiente.fmach.it/
--
Regards,
Doc
--------------------------------------------------------------------
Vincent B. Robinson |
Department of Geography | E-Mail: [email protected]
University of Toronto | Voice: (905)828-5299
at Mississauga | Fax: (905)828-5273
Mississauga, Ontario |
Canada L5L 1C6 |
_______________________________________________
grass-user mailing list
[email protected]
http://lists.osgeo.org/mailman/listinfo/grass-user
