Sen2res plugin on SNAP 9 and Bad Results

[dpoursanidis]

Dear community,

We constantly use the plugin for bands downscaling instead of simple resamplings.
After SNAP 9 and Sen2Res ver 1.1. and the use of L2 Sentinel 2 products with the new folder structure, the results on B1 are bad visually but also quantitatively.

and values in comparison to a Matlab based approach by Lanaras

Has anybody phase similar situation?

Since i have a batch of imagery which use the older folder structure and are L1B products, the process using same setup works smoothly.

Any help, update, idea, workaround solution is much appreciated.

Looking forward

Dimitris

[dpoursanidis]

I would like to add here that when using S2 data with the previous folder structure and xml, all gone smooth and fine!

[oana_hogoiu]

Hello,
Unfortunatelly, Sen2res is a third party plugin, not maintained by SNAP team.
It was developed by @nicolas.brodu, you can see below.
There is no newer version of this plugin that could take into account the new Sentinel-2 format products.

Sentinel-2 Super-Resolution (all bands at 10m): Snap plugin now available!

I have a simple question Sir… I am trying to make super-resolution Sentinel-2 imagery. And by using your Sen2res, it was possible to create all bands 10m resolution images. And then, I made RGB and NIR bands to 5m resolution by using Deep Neural Network. Then, is it possible to use these 5m bands again to create other 10m bands to 5m enhanced resolution? Many thanks!!

[Jan]

@dpoursanidis Hello

Can I please ask what you mean by “…the new folder structure”?

Thanks

Jan

Jan Jackson

OPT-MPC S2 Technical Manager

[dpoursanidis]

Hi

I mean the new format of S2 which has in the granule the 3 folders 10-20-60 instead of the previous where all bands where in that folder only.

Best

D.

[dpoursanidis]

Hi

I know that is a third party plugin which does amazing qualitative work. However, Nicolas does not maintain the plugin following the new developments of ESA on the product format, hence now is useless, unless ESA wanna make it work by incorporating in its standard tool.

D.

[dpoursanidis]

I notice that L1C has the structure Granule>Bands inside but L2A has the structure with the folders, hence when we need to use that product for downscale, the tool does not work properly.

[Lantzanakis]

Hi @Jan ,

There is a new band inside R20m folder, B01_20m.

Sen2Res doesn’t work properly for B01 so i guess this is the reason.

Also there are changes in metadata structure (.xml) for the conversion from DN to BOA Reflectance

But Sen2Res seems updated of this change