Create C2 matrix using Setninel - 1 data and Sentinel - 1 Toolbox

Microwave Toolbox Polarimetry
#17

Cazals, I’m not sure I see your problem. Did you use the Polarimetric Matrix operator and select C2?

#18

Thank you for your help.
Yes : first I debust the Sentienl-1 SLC data, then I run the “Polarimetric matrix generation” selection C2 matrix. I get the following error message :
A problem occured during the target product initialisation.
Type : OperatorException
Message : Polarimetric-Matrices : C2 cannot be generated for this product

The Polarimetric decomposition is running without problem but this is not what I want, I want the covrariance matrix (C2).

#19

Which product is it? Is it dual pol?

#20

Yes, it’s an SLC Sentinel-1 Dual pol VV/VH (S1A_IW_SLC__1SDV_20151030T061552_20151030T061619_008378_00BD4A_6804).

#21

I have tried the same procedure as proposed

  1. Input: S1A_IW_SLC__1SDV_20170331T055916_20170331T055944_015932_01A445_4104.zip
  2. TOPSAR deburst
  3. Matrix generation
    In matrix generation I get the following error:

A problem occurred during the target product initialisation
Type: Operator Exception
Message:Input should be a polarimetric product

I skipped that matrix generation step according to above suggestions and again I get same error in Decomposition step

  1. Decomposition

A problem occurred during the target product initialisation
Type: Operator Exception
Message:Input should be a polarimetric product

Any suggestions ?

#22

Sentinel-1 data is dual-polarized by definition.

To derive polarimetric signatures you mostly need quad-polarized data. There is a dual-polarized decomposition but it delivers different results.

#23

I tried to create C2 matrix from sentinel-1 SLC IW data. my system has 16GB RAM and corei7 processor.when I try to preprocess VV and VH at the same time JAVA heap space error appears. In order to solve this problem firstly I used S-1 Tops Split and then my preprocessing step is :
1- apply orbit file
2- radiometric calibration
3- debrust
4- polarimetix matrix generation (C2)
but in this step this error appears :
A problem occurred during the target product initialisation
Type: Operator Exception
Message:Input should be a polarimetric product

#24

If you calibrate the product, please select the complex output option.
You cannot perform polarimetric tools on Sigma0.

#25

@ABraun I exported complex version and got result. thanks

#26

Hi @ABraun I want to have sigma0 from C2 matrix elements of sentinel-1 SLC data . do my processing steps true?
1- S1 tops split
2-apply orbit file
3-calibrate
4-deburst
5-merge
6-matrix generation(C2)
7- multilook
8-speckle filter
9- terrain correction

#27

Yes, that looks correct. Multi-looking is optional, the data is resampled during RD Terrain Correction anyway.

1 Like
#28

The value of C11 is between 0 to 3 and the value of C22 is between 0 to 13. does it seems correct?

#29

@ABraun could you please help me for this question?

#30

patience please - sometimes we spend our weekends outside this forum.

If we look at the formula, we see that C11 and C22 are calculated like this:
C11 = |VV|²
C22 = |VH|²

C2error

To result in the reporded range of values (you are probably looking at the 95% range from the color manipulation tab*), this would mean that
VV ranges between 0 and 1.7
VH ranges between 0 and 0.36

To me this sounds quite logic.

  • You can compare the entire range of values in the statistics tool
    grafik
#31

@ABraun thanks for your answer
Do you mean that my my answer quite logic or your range of VV and VH ? and If my answer is not logic what would you suggest to correct it?
thanks in advance

#32

which answer do you mean?

#33

you said the range of VV between 0 to 1.7 and the range of VH between 0 and 0.36

#34

I mean my range of C11 was between 0 to 3 and C22 between 0 to 13

#35

sorry, I have read 0.13 instead of 13.
I don’t see a reason why the ranges should not be correct.

#36

Thanks for your response