Jump to content


supervised classification - forest layer

- - - - -

  • Please log in to reply
No replies to this topic



    Key Contributor

  • Validated Member
  • PipPipPip
  • 96 posts
  • Gender:Male
  • Location:Cracow
  • Poland

To the mountain project I'm working on I need to do a supervised classification and get one layer - forest.

I have Ilwis http://52north.org/i...r...&Itemid=127 to disposition and pretty Landsat ETM+ scenes http://landsat.usgs....ery_Release.pdf in the end of the month.

I'm quite new to teledetecion have a small academic experience and before next steps in Ilwis and literatures I would like to ask some question which posiblly save me some time, so..
1. Is geometric correction needed for such new scenes like ETM + are ? I guess no
2. Is radiometric correction needed to correct spectral answers of the shadowed pixels..it's a mountain area, I guess yes but maybe for such a homogenic layer as forest is it's not necessary?
3.Which band should be the best to classify forest layer? I guess infrared?
4. Do you know any FOSS except ILWIS which could do classification, corections etc. and it's worth a try?
5. Any trick, tips, workflows, links to tutorial with forest classification from people that did that what I'm going to before?

Thanks in advance for any help

0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users