Level-1 processing:

• data import and metadata homogenization (compatibility: Landsat-5/7/8, Sentinel-2A/2B)

• equalization of acquisition- and illumination geometry

• atmospheric correction (using SICOR)

• correction of geometric errors (using AROSICS)

Level-2 processing:

• spatial homogenization

• spectral homogenization (using SpecHomo)

• estimation of accuracy layers

=> application oriented analysis dataset

Usage via WebApp

The recommended way to use gms_preprocessing is to setup the WebApp (see the gms-vis repository) providing a UI for GeoMultiSens. Using this UI, existing satellite data can be explored, filtered and selected for processing. New data homogenization jobs (using gms_preprocessing) can be defined and started. All configuration parameters of gms_preprocessing are accessible in the UI.

Usage via console interface

Homogenization jobs can also be created and started using the command line interface. Documentation can be found here.

Here is a small example:

# start the job with the ID 123456 and override default configuration with the given one.
>>> run_gms.py jobid 123456 --json_config /path/to/my/config.json

There is a default configuration file, called options_default.json. This file contains the documentation for all the available configuration parameters.

Usage via Python API

There is also a Python API that allows to setup and start homogenization jobs by a Python function call.

This is an example:

from gms_preprocessing import ProcessController

configuration = dict(
    db_host='localhost',
    CPUs=20
    )

PC = ProcessController(job_ID=123456, **configuration)
PC.run_all_processors()

Possible configuration arguments can be found here.