Hello, What I am trying to do is create an exercise for my students to test them on their supervised classifications. I have them making their own training ROI's for Landsat TM data. What I need is the ground truth data, in this case in the form of Land Use maps provided by the county. My process thus far is: Take Landsat image, clip it by the Land Use map, and give it to the students to classify. I took the Land Use map, saved it as an ENVI file, changed the header to Classification, set the classes and went to use the confusion matrix with ground truth image. I get the error "The Class Image and Ground Truth Image Do Not Overlap.". I know this is not true because they are in the same projection and one was used to clip the other. Is there a better method for making ground truth maps? I.E. That produces usable ground truth maps?