Processing tool package written in bash and python for image correlation of coregistered SAR images using the AMES stereo toolbox (https://stereopipeline.readthedocs.io/en/latest/). The coregistration is done using the AMSTer toolbox(https://github.com/AMSTerUsers/AMSTer_Distribution/).\ The processing chain additionally compute the time series analysis with NSBAS .

1. AMSTer Coregistration

ALL2GIF.sh YYYYMMDD /full-path-to-LaunchParam.txt 100 100

Launch command in AMSTer/SAR_CSL folder, coregistrated images stored in AMSTer/SAR_SM/AMPLITUDE folder.

2. Initialize ASP-SAR

amster2aspsar.py /full-path/AMSTer/SAR_SM/AMPLITUDES/SAT/TRK/REGION /ASP-SAR/working-dir [--s1]

3. Generate pair network

For example, in /ASP-SAR/working-dir/PAIRS

lns_All_Img.sh CSL-dir pair-dir

Prepa_MSBAS.sh pair-dir Bpmax Btmax

Eventually use Delaunay triangulation:

DelaunayTable.sh -Ratio=r -BpMax=bp -BtMax=bt

You can copy your table of choice as table_pairs.txt.

Display the network:

amster_plot_network.py table.txt

Add a new pair to the table:

amster_table_add.py table.txt d1-d2 d2-d3 ...

Remove a pair from the table:

• Open the table in VI
• Cursor on the line to delete and dd

4. Image Correlation using ASP (through aspeo)

Generate the parameter file:

aspeo new preset

Update the parameters of interest

Launch the stereo process:

aspeo pt aspeo.toml -v

5. Export the results for NSBAS

aspsar2nsbas.py asp-sar-dir

6. Run TS inversion in WORKING_DIR/NSBAS_PROCESS/MASKED/H|V

Update the parameters of inversion: input_inv_send

Launch the inversion:

invers_pixel < input_inv_send

7. Geocode results

You can create a GEOCODE directory to store geocoded results.

Choose one of the date directory to host the geocoding steps and check the geocoding parameters in LaunchMTparams.txt.

Note that the sm run must have corrected filepaths. Run RenamePathAfterMove_in_SAR_SM_AMPLITUDES.sh SAT if not the case.

am_geocode.py file --amster=sm/amplitude/trk/region/one-of-the-dates/ --outdir=../../GEOCODE/[HV]

If conflicts of python version, you can run once am_geocode, then in the sm/amplitude/date dir run ReGeocode_AmpliSeries.sh /full-path-to-params, then run am_geocode again to get back the results.

GDAL mutli-band rasters will be split by band. To reconstruct the cube, check merge_star.py.