Rss

Sen2Res

Sen2Res is a processor to enhance the spatial resolution of Sentinel-2 products till 10m/pixel preserving their reflectance.
Sentinel-2 MSI does not have a panchromatic band, but it contains four bands at 10m/pixel. Sen2Res works by building a model of how information is shared between these bands (i.e. a pixel content, independant of the reflectance), and what information is specific to these bands (i.e. the color of that pixel content). The model is then applied to unmix the 20m/pixel and 60m/pixel bands while preserving their reflectance.

Sen2Res can be quite computationally intense. It is recommended to try it on a small Region Of Interest (ROI) first, then increase the size of that ROI according to the computational power at your disposal.
For any questions about the usage of Sen2Res, please follow the general STEP forum area dedicated to Sen2Res: http://forum.step.esa.int/t/sentinel-2-super-resolution-all-bands-at-10m-snap-plugin-now-available/6927.

Sen2Res is provided as a plugin for SNAP version 5 or higher and it works for Linux, Windows 64 and Windows 32.

Steps to install Sen2Res:

  • Download the sen2res-1.0.nbm file from the link
  • Update SNAP (Menu -> Tools -> Plugins -> Updates -> Check for updates…)
  • Install Sen2Res module:
    • Menu -> Tools -> Plugins -> Downloaded -> Add Plugins.. (select the .nbm file)
    • Click on Install button
    • Next, check I accept the terms… and click Install
  • Restart SNAP (mandatory)

Latest release:

  • 1.0: It was released on October 12, 2017. This version runs on Linux and Windows 32 or 64 bits: sen2res-1.0.nbm

Details on the method, together with the (open) source code, are provided at http://nicolas.brodu.net/recherche/sen2res/. You are welcome to reuse the plugin framework as a basis for implementing other resolution enhancement methods. To recompile from source, type mvn install package from the source directory. The plugin should be generated in a target subdirectory.

Sen2Res was developed by Nicolas Brodu working at Inria as part of the Geostat project.

Enjoy!