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

Thanks, i will try it again.

Hello to all step forum members
I processed all of my images successfully and exported them to stamps. i ran stamps1-6 without any problems. but unfortunately my results are strange as below . it will be very appreciated if help me to solve this problem.

Hello, I have the same error, could you solve it?

Good afternoon, I have a problem executing mt_prep_snap in ubuntu 20.04, I am already re-related to the stamps working process and I have never had this problem, in ubuntu 18.04, only I changed computer and I had to redo everything again. I have installed the two versions of stamps and on both I get that error.
I think I have installed everything correctly but I get that error, please someone could help me. Beforehand thank you very much.

segmentation fault error with mt_prep_snap is due to the gcc version employed to compile the stamps/src folder.

this normally happens when using the provided gcc in Ubuntu 20.04. With version 18.04 does not happen.

1 Like

You do not need to change the operating system to solve the segmentation fault error. This issue can be solved on Ubuntu 20.04 by adding the old xenial repositories to /etc/apt/source.list:
##GCC 4.8
deb http://dk.archive.ubuntu.com/ubuntu/ xenial main
deb http://dk.archive.ubuntu.com/ubuntu/ xenial universe
Then:
sudo apt-get update
sudo apt-get install gcc-4.8 g++-4.8

Finally in the make file PATH_TO_STAMP/src/Makefile change line 40 from CC = gcc to CC=g++-4.8

The same thing can be done in Debian 10 by adding deb https://deb.debian.org/debian/ jessie main to /etc/apt/source.list and then do the same steps as in the Ubuntu case. This will add Debian 9 Jessie repository to the list.

I hope this can help.

5 Likes

thank you for sharing this solution. This will surely help others because this problem comes up quite frequently.

1 Like

Hi!

I just want to ask if you have resolved this Warning? If not, I want to ask the opinion of @ABraun, @mdelgado, and @thho.

Thank you.

As it states, the message just informs you that a parameter was automatically changed to allow the further processing of the script.

1 Like

There is a bug in some programs used by mt_prep_snap. Apply this patch to fix it. I submitted the patch to the development repository, but nobody replied.

BTW, the fault here is not with gcc, but with the program. Some change in gcc just exposed the mistake in the program.

3 Likes

Can you share here the files you have modified?

They are in the pull request. Can you see them?

Anyway, you just have to change the return type from int to void for functions cshortswap and similar in all the files where they appear. In C and C++ it is forbidden to not explicitly return something in a function whose return type is not void.

2 Likes

Have you solved this problem yet, would you please tell me what is the problem?

I have been testing minimum surfaces but they have all given me the same error, would you know roughly tell me what could be the minimum dimensions in which good results are obtained in the detection of PS? @ABraun

I don’t really know if there is a lower size limit. What areas did you test?

I did not obtain a satisfactory result in any case, perhaps due to having started from too small surfaces; the last one used was 500m x 500m; so stop dwelling on it. and I continued using large surfaces; If I will try the minimum surface test in the future again, I will let you know

1 Like

Hi modeste

Please how Can I get this figure ??

Thanks.

Regarding the “Dimensions of arrays being concatenated are not consistent” error, please use the search function of the forum, this problem has been discussed and answered multiple times:

1 Like

Dear Fillppo
I have the same problem as you,when I run the stamps(3,3) , I have 9 PATCH, but some PATCH show this:
Warning: Not enough random phase pixels to set gamma threshold - using default threshold of 0.3

In ps select at 184
In stamps at 357
In this case, there are still PS points.

But in some PATCH ,show this:
PS_SELECT: Initial gamma threshold: 1.000 at D_A=0.05 to 1.000 at D_A=0.40
PS_SELECT: 0 PS selected initially
PS_SELECT: Reestimation gamma threshold: 1.000 at D_A=0.05 to 1.000 at D_A=0.40
PS_SELECT: 0 PS selected after re-estimation of coherence

I have no idea to save the problem, can you give me some advice?
thank you, I am waiting for your reply!

there seem to be patches which have a very little amount of PS, which can have two reasons

  • there water bodies in larger parts of your data
  • the patches cover small areas., then you could reduce the number of patches in the mt_prep command