Which are the most common approaches for pre-processing SAR & InSAR Sentinel-1 Data for snow and ice applications?

Greetings,

I'm about to start my Msc thesis and I've been told to use the MATLAB signal processing toolbox to do the pre-processing. However from what I can find online, the standard way is to use one of several available softwares (SNAP, GAMMA, ..) to do this type of work and not MATLAB. The point of this post is to ask the community which methods they use and if they use MATLAB at all and to what extent? The organisation I'm doing this thesis for wishes that I in the end attempt to automate the process and from that POV MATLAB seems like a good candidate.

After the pre-processing I need to do image classification in ArcGIS Pro, so that may also be a consideration. Also, I have the R2023b release via my university, but the computers at my campus haven't been updated to the latest version, that might also be a consideration.

Thank you for all input,

/Jakob

Answer 1

[Shivansh]

Hi Jakob!

It seems like you want to analyse the usability of MATLAB for your thesis. You can use MATLAB for pre-processing SAR (Synthetic Aperture Radar) and InSAR (Interferometric SAR) Sentinel-1 data for snow and ice applications.

Some common approaches for the pre-processing of SAR data in MATLAB can be:

1. Speckle Filtering: SAR images can sometimes be inherently affected by speckle noise.MATLAB has built-in functions for speckle reduction, such as "wiener2" for 2-D Wiener filtering.
2. Terrain Correction: Terrain correction can be performed using the SAR Simulation Terrain Correction (SARSim-Terrain Correction) method. MATLAB's Mapping Toolbox could be used to assist with geospatial transformations.

There are many builtin functions in Signal processing toolbox which can be useful for preprocessing purposes for SAR dataset.

MATLAB also has its benefits with the automation of these processes because of the scripting, batch processing and wide support of toolboxes. You can also use MATLAB's App Designer to create a graphical user interface (GUI) for your processing pipeline.

For the integration with ArcGIS Pro, you can export the processed images from MATLAB in a geospatial format that ArcGIS can import, such as GeoTIFF. MATLAB's Mapping Toolbox allows you to work with and export geospatial data.

As for the different MATLAB versions, while the latest features and functions are available in R2023b, you can check for the feature compatibility for a particular release from the MATLAB official documentation.

You can refer to the following documentation links to get a better idea:

1. Signal Processing toolbox: https://www.mathworks.com/help/signal/index.html.
2. wiener2: https://in.mathworks.com/help/images/ref/wiener2.html.
3. SAR examples: https://www.mathworks.com/help/radar/synthetic-aperture-radar-sar.html.

I hope it helps!

[Jakob Papirov]

Hi Shivansh,

Apologies for not replying sooner, I'll make sure to reply on Thursday. Thank you!