Classification Lab.
In this lab you will practice creating a classification of images using 3 different
methods.
Part 2- Supervised Classification
- You will perform a supervised classification of your "field" area or if
you don't have one yet (you should !) use
the copy you made of R:\satellite\Virginia\TM16_34_1999_1-7_shen_valley_clip.ers
The
process is illustrated here.
using the following classes
- urban/transportation
- pasture
- deciduous
- pine
- water
- "other"
If you are using a new area, come up with you own list and run it by me
before you proceed. You can use raw bands, or virtual datasets containing
ratios, principal components
or
any combination
of
the three
to develop
the classification, but you should make your training regions over a
rgb321 or rgb432 image so that you can see what you're marking off. .
- Export the classification to ArcMap using the geolocation process that
you completed last week.
For this lab, create
- some views of your supervised classification training regions
- one screenshot of how the training regions look on the scattergrams from
one IR band and one visible band.
- your supervised classification map
and an evaluation of how well it does it's job.