Hello everyone! I'm facing a problem that I've been stuck on for almost a week... My aim is to map the environment and, to do this, I'm carrying out supervised classifications in order to distinguish the different elements of the landscape. I need to carry out these classifications on all the bands of my satellite images (Sentinel-2) at two different resolutions (10 and 20 meters) and for 6 different dates, in order to determine which are the most likely to give a coherent result for the final mapping. As the number of bands to be classified was very high, I wanted to automate the supervised classification process. So I tried lots of different things, none of which produced a satisfactory result. I wondered whether it was really possible to automate this, and if so, how? The idea would be to have a single Training Samples valid for the different bands (perhaps taking into account only georeferenced points to be exported as Training Samples?).
Thanks in advance!
