Good luck. You can share your results in here with screenshots and we can maybe help again if you are stuck somewhere.
I am having my training data sets as point file. Is it possible to classify S1 data in SNAP using points?
I don’t think so. In any case for SAR data using a single pixel would be bad.
If I have the polygons, is it possible to do supervised classification in S1 images with SNAP?
yes. Open your SAR image, menuy > import > other > shapefile
Then use the polygons as training vectors in the classification modules by selecting the column that has the class information.
How do you execute a Support Vector Machines classification to map rice areas in Sentinel-1 toolbox? thanks before
SVM is currently not implemented in SNAP
Thanks for the reply ABraun. Do you have any suggestions on what classification method i can use to map rice areas that can be achieved with SNAP Sentinel-1 toolbox? My data is IW GRDH by the way
the random forest classifier is quite good if you have data of different scales. But it needs more than just 2 rasters.
For the case you only have VV and VH you can use the KD Tree KNN classifier.
Okay Abraun how do you execute a KD Tree KNN classifier on SNAP? I’ve tried looking at http://step.esa.int/main/doc/tutorials/ and the ESA training courses but couldn’t find what was needed. Sorry for the many questions.
collect sample areas and use them for training.
Please see here:
Thanks a lot ABraun one last question how do you digitize polygons in SNAP? Is it the same as in Arcmap? how about points?
points are called pins in SNAP
you can digitize them with this toolbar, it is quite intuitive
can the enmap be used on sentinel-1 data classification?
Is it now implemented`?
I’m afraid no.
I have version 6 but I did not find it and I did not know about version 7 but it seems it is not implemented there. Thanks for answer.
Did you not receive a notice about version 7 every time you check for updates? Anyway, version 8 is some weeks away not but you should try 7 too.
Thanks I will try it but Do you know it has SVM classifier or not?