Hello,
after preprocessing of S1 GRD image in SNAP, and export in GeoTIFF format, I want to perform a Supervised Classification in SAGA. I import S1.tiff image and samples as polygons and when I want to do a classification I get this error
Is there some step I’ve missed? I have also tried to export S1 image in ENVI format, got the same error. Any help would be appreciated.
SAGA should be able to directly open the BEAM DIMAP format, so the conversion to tiff is not necessary.
@ABraun I’ve tried that, but when I drag and drop .dim file into SAGA, it does not respond (import the image).
However I’ve solved it by importing it in .img format and then saving to SAGA format, but still want to avoid this step… So You can open normally .dim file in SAGA?
Thank You
drag-and-drop is never a good idea because the program doesn’t know what it gets. I would always recommend the import function to guarantee a proper display and use of the data.
Why saving it in SAGA format at all?
@ABraun I’ve tried drag-and-drop because under Menu->Import in SAGA it doesn’t show .dim extensions as Recognised files so I didn’t know what to do else…
the dim is just a wrapper for the ENVI rasters inside the data folder (img+hdr)
@ABraun thank You very much, everything is successfully importing in SAGA GIS
It’s not the first time I read you saying that. I agree that .img/.hdr files can be easily read by any GIS software.
However, I’m not quite sure how to properly import the .dim/.data into QGIS. I’ve found the plugin Processing BEAM and SNAP algorithm Provider but I doesn’t really help me out.
My brain should have skipped an elementary step but if you can recommend me a procedure in QGIS to import the bands and metadata of BEAM-DIMAP format, I would be very grateful
What kind of metadata do you need in QGIS? Regarding projection, pixel size ect, I never had problems in QGIS. But yes, things like acquisition date, tie-point grids or sensor type are not automatically recognized by QGIS. But on the other side, I don’t know any process which could make use of such information, to be honest. Or did I miss your point?
If more rasters are within a stack, I mostly use gdalbuiltvrt to create a virtual stack, without having to (physically) write a new one.
Let’s say I specifically need the geometryDcPolynomial present in the acquisition metadata, which can be useful in some applications. I don’t really know how to use it easily out of SNAP.
I have no solution for that, sorry. My approaches were probably less advanced, so I was fine with the raster information 
Would it be worth reading the dim file with python?
That’s currently what I’m doing, but out of QGIS 
Still thank you for the answer