David Renna

I am ready for an exciting semester.

Tuesday, July 20, 2010

Module 4: Supervised Classification Challenge

This week we were required to take an image of Germantown, Maryland and do a Supervised Classification in ERDAS Imagine of the image to known land cover types. We were required to use 14 categories and label each type. Then using histogram values, we were to categorize the land in the image.
I really did not have too many problems with the assignment. I was able, through trial and error and the known value ranges, to match all but two categories. The grass areas and Ag 4 areas had a bit high histogram values. I found that if I adjusted these values many other categories fell out of range so I felt that this was as close as I could get. I did spend many hours adjusting the Euclidean Distance values to get the histogram values within range.

Week4:Supervised Classification

No comments:

Post a Comment