I think it doesn’t work properly.
Can you please help me in this.
25000 lines is quite huge.
It is recommended to set some overlap (200 pixels) for such large files. Delete all temporary files, modify the config file (maybe also set MST instead of MCF) and run the command again.
1 Like
Shall I have to repeat all the procedure again?
you can create a subset from the interferogram to reduce the amount of data (if possible), but this did not cause the error.
You have to repeat the unwapping anyway because it did not finish successfully.
Thanks, but I am not able to subset from the interferogram.
Is there any other method to reduce the data size??
depending on the type of deformation, multi-looking might be an option. How does your interferogram look like?
20x20 m will still allow most analyses and significantly reduces the file size.
But as I said, the error message was propably not caused by the size.
I completelly repeat the process and also reduce te size of my file. but it still not working. Its took too much time.
your data is almost completely decorrelated. I’m not sure if you can retrieve valid results from such an interferogram. If there is noise in it, the unwrapping will not make it better and rather produce random patterns.
Is there much vegetation in your study area? Did you select the best possible image pair?
For comparison, this is how an interferogram looks like in the best case:
credits to @qglaude
1 Like
Ahah thanks for the credits.
@Tamanna Phase unwrapping is long (can be few dozens of minutes up to several hours). Coherence map is key to estimte in advance if the phase information is usable. But on top of that, don’t forget to multilook and filter your interferogram (goldstein phase filtering) before exporting to snaphu. It is crucial if you want a usable phase information. The first operation increase the phase quality by estimating by aggregating neighbor pixels and reduce drastically the size of the image (using 6x2 multilooking, your image goes from 25000x16000 to 4000x8000, reducing the amount of info by a factor of 12). Depending on your case, you can even increase this parameter. The filter operation enhance the interferometric signal by reducing its noise while preserving its pattern. It ensures that between 2 pixels, the \pi shift criteria is better respected.
To summarize, before unwrapping :
- Check coherence
- Multilook
- Goldstein Phase Filter
Hope it helps
1 Like
Actually I am working on Himalayan terrain to estimate the snow depth by taking a pair of non-snow and snow cover data. There is very less amount of vegetation is present.
I didn’t used multilook in my process. Does it effect my result??
It does. Both in computation time and results’ quality.
Hey, I took a new area and create an interferogram
but after unwrapping I tried to import it and it shows this error
what does it mean?? can you please tell me.
was the unw_*img file created correctly and which file did you select for the input?
yes i complet the unwrapping porcess successfuly.
so which file did you select for the snaphu import?
Please also check here: Problem in Snaphu Import section
thanks for giving your precious time to me.
Actually when I apply "goldstein phase filtering " on data after DInSAR >multilook>goldstein phase filtering. then it generates two phase in my bands. which make complication in export.
when exporting this file it takes “phase_ifg_srd_vv” not the other one in snaphu.conf file.
I think it should read the “phase_ifg_vv”. See the difference in both file
can you please tell me, where I am making mistake.
very strange. I haven’t seen this behaviour, but maybe it is caused by the topographic phase. Can you please go back to the interferogram formation and directly activate topographic phase removal here:
Then you can directly proceed with debursting, multi-looking and filtering.
