Hi, After classifying an image, it has different areas in one class.
In the picture, in class 9 with red color. it can be seen that the arable land areas are in one class with urbanism.
Is it possible to divide a class and then merge it into another?
Thank you for help Petr.
No, splitting or merging classes afterwards is not possible.
You might need to refine the classification parameters or use another classification method. Which one have you used?
This one is unsupervise classification( K-Means Cluster). I try maximum likehood classification but for this class final result was almost similar. When i viewed areas in spectral wiev, they have almost the same values.
you could also use the Mask Manger to create sub-classes based on manual thresholds. For example
IF raster == class1 AND IF b2 > 0.5 THEN class1a ELSE class1b
Yes, indeed this is possible. But updating the classes and having a new index coding is not possible.
Would be a nice feature, creating an index coding manually.