Coregistering of SAR images

s1tbx
#1

Hi guys,
I’m having real trouble coregistering two SAR images from a region in Svalbard, Norway. What I’m trying to do is derive surface glacier velocities from Sentinel-1b GRD IW product pairs. For each of the two images, I created a ROI subset, then applied the orbit file, removed the thermal noise and calibrated. Then, when trying to coregister (DEM-assisted-coregistration) the two images (and subset them after that - but the error already occurs in the coregistering step because I tried that separately too) the process runs and executes but the resulting product somehow just doesn’t seem to exist. It just shows a dark grey background.
I tried it with different DEMs suggested in snap at that step and also tried it with an external DEM from the Svalbard region but the result would always be the same (aka none…).
Does anyone have an idea what I am doing wrong??
Thanks a lot for your help!!
Kira

#2

Can you paste the image filenames ?

#3

Yes, of course!
It’s

  • S1B_IW_GRDH_1SDH_20171113T154443_20171113T154508_008265_00E9E4_FD12
  • S1B_IW_GRDH_1SDH_20171125T154424_20171125T154449_008440_00EF38_43FE
  • S1B_IW_GRDH_1SDH_20171125T154449_20171125T154514_008440_00EF38_041D
    but I assembled the two latter with slice assembly before subsetting them to my ROI.
    I am only using the HH polarization
#4

Well … I’m already stuck at the DEM-assisted coregistration step. I don’t know why but I’m stuck at 0% of the process without any error.

Nevertheless, when I do the DEM-assisted coregistration on images of my area of interest in Antarctica it works. I can’t say why. Maybe the today’s connection issues do not allow me to download the ACE30 data?

#5

Does it maybe create the output although the process seems to not be finished?! It did that to me too a few times…
So you think the problem lies in the ROI or the DEM of that ROI?

#6

Is it maybe possible to skip the coregistration step? If I subset the original images to my ROI using geographic coordinates before doing the preprocessing (orbit file, thermal noise removal and calibration), the images should already be coregistered right?

If so, how do I combine the images then into one file in order to be able to subset them into master and slave in the next step?

#7

No, it is not… I tried mosaicing the two images but when trying to derive surface offset (Radar -> SAR Applications -> Offset Tracking) this error occurs: “A problem occurred during the target product initialisation. Type: OperatorException. Message: Source product should be a coregistered stack created by DEM-Assisted Coregistration.”
… so it doesn’t seem possible to skip the coregistration step although the images should already be coregistered…

Does anyone have an idea on how to solve this??
I would be super happy about some input!!

Thanks in advance!

#8

Yes it creates the output but it’s almost empty. SNAP creates the placeholder with metadata and geometry data.
You can open the “empty” image in SNAP but it’s pointless you will see nothing.

Again, my guess is that connection issues may stop the DEM-assisted coregistration process. Not sure tho

And no, you cannot avoid the coresgistration step

#9

hi， have you solve this problem? i meet same question too