for supervised methods, try i.smap. A comparison of SMAP with maximum
likelihood and ECHO (extractions and classification of homogeneous
objects) may be found at:
http://soils.ecn.purdue.edu/~mccauley/www/abstracts/smapecho-abs.html
(to appear in IEEE Trans. Geo. & Remote Sensing). Most of this work, with
the exception of ECHO, was done using GRASS. (ECHO was performed using
MultiSpec, Mac software. For more info, see:
http://dynamo.ecn.purdue.edu/~biehl/MultiSpec/
)
--Darrell
Milton Ribeiro (mribeiro@carpa.ciagri.usp.br) writes on 17 November 1995:
I have interest in references & literatures of Scene segmentation using
(un)supervised methods. I'm a newbie with GRASS and I need to do
comparations of theses methods in a perturbed (forest, regrowth,
deforestations) site.I'm very thanks for all help!
Miltinho (RIBEIRO, M.C.)
Quantitative Methods Laboratory
Forest Science Department
ESALQ/University of Sao Paulo
Brazil
--
Darrell McCauley, PhD; mccauley@mcs.com; http://www.mcs.com/~mccauley/