I have some limited knowledge regarding ENVI. I also used only ENVI classic before.
Using 5.2, I have several hyperspectral images. I opened them all in ENVI 5.2 (it's nice compared to classic where you have to open a new display for every single one).
Took me a while to figure, that when I use ROI, I must also select the image I want the ROI to display in.
Anyway, my question is, once I create a set of ROI from every image (some images might have 2 of them), how can I "combine" them?
Basically I need 2 sets of ROI's, one for "good" roads, and one for "bad" roads, coming from several hypespectral images. I mapped these roads using ROI, and will be using either Min/Max classifier or Support Vector Machine. However, as mentioned the ROI's are under each image. When I do the classification, I want to use the spectral signature of all my ROI's which I made for "good" roads, then all the signatures of my ROI's I made for "bad" roads.
I also assume I need a 3rd ROI set for "others" which will include houses, trees, lakes, etc.
Thanks!
EDIT: I just realise I might have to "tile" the images I work on. Basically take all 8 .img files and make 1 big one (ENVI 5.2 looks like it does that when I have the ability to open all 8 in one display, but I guess they are still very much separate).
Is there a tutorial available how to "combine" into 1 .img file if I have to do so?
Cheers!
|