To begin this project, I defined training sites based on the Nation Agricultural Image
Program (NAIP) 2006 photo. Once defined, it was then exported to Idrisi, where I ran a MAKESIG,
in order to create a
signature group. Then with that signature, I ran a classification
function. I ran a MAXLIKE, PIPED, FISHER, and a MINDIST classification to
see which one portrayed the information the best. While there were
advantages and disadvantages to each
classification technique, I ended up choosing the MINDIST classification
because I felt as though it displayed the information the most accurately.
Next I imported the new map into ArcMap, along with the
provided 2001 Land Use image. I then ran a RECLASS on both of the maps, in order to give them the same Land use
categories and colors.
Lastly, I imported both maps back into Idrisi in order to run
a Change Analysis. I chose to use Crosstabulation to check the
accuracy of our assessment. Finally, I ran ASSIGN in order to classify the crosstab analysis as either change or
no change.
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.