[GRASSLIST:10040] How to sample single pixels as one training class?

Anja Drews anja.drews at uni-oldenburg.de
Mon Jan 30 06:17:41 EST 2006


Dear list,

I am trying to put a training data set together for maximum likelyhood 
classification.
The training site data for one class come from single pixels distributed 
over the whole image.
I imported  this sites with v.in.ascii, did v.to.rast, and wanted to run 
i.gensig for the generation of the image statistic and the signiture 
file. My 25 distributed pixels for one training class are considered at 
25 individual classes.
How can I avoid this? All 25 pixels belong to one class.

Kind regards,
Anja




More information about the grass-user mailing list