Definition Slave and Master Images in a set of images

Hello Dear Friends,
I am analyzing the two Sentinel 1A images taken from two different times for landslide detection. However, I am curious that the two images will be enough to detect the landslide with regard to accuracy and reliability. If not, I want to use for instance 10 images (e.g. 9 slaves, 1 master). If possible to use 10 images for a better and accurate result, how to define the slaves and master images in SNAP software. I mean in which step should I define the slaves and master images.

It depends on the approach you’re following to reach your goal,


First of all, thanks a lot for your response FalahFakhri, I did the same thing and I use 4 images taken at different times. I used S-1 Back Geocoding to stack a set of images. Then I proceed the processes with several other steps like ESD -->Interferogram–>TOPSAR Deburst–> TopoPhaseRemoval–>Multilook–>GoldsteinPhaseFiltering–>SnaphuExport(on Cywgin) --> SnaphuImport–>PhasetoDisplacement. However, it gives the error in the picture below in Snaphu Import Step. I repeated these procedure several times, it gives the same error again and again.

I think the snaphu could read only one phase, and in your case, there are many phases, in case you’d like to create stacking average so this approach you have followed is not recommended at the moment, Also when you did export snaphu, Do you have any idea which phase is exported?

I did some effort and I discussed with other colleague in STEP FORUM @ABraun you could find more about the stacking method here,

1 Like

Thanks again @falahfakhri, I will check the discussion with @ABraun about this issue. By the way, you imply that two image analysis depends on my approach. In general, what do you think about the analysis of just two images for landslides detection with respect to reliability? My approach is this:
Split–ApplyOrbit–ESD --Interferogram–TOPSAR Deburst– TopoPhaseRemoval–Multilook–GoldsteinPhaseFiltering–SnaphuExport(on Cygwin) – SnaphuImport(on Cygwin)–PhaseToDisplacement–RangeDopllerCorrection.

I meant it depends on the technique you’d like to follow, in case of PSI, you’d need at least 12-20 images, however following the DInSAR technique at least two images should be processed, and so on concerning the other techniques,

I think preparing and processing series of S-1 and ENVISAT ERS… have been mentioned for example
under this topic, you could find a lot of things

How to prepare Sentinel-1 images stack for PSI/SBAS in SNAP 5

I found this, I think it is important to you,