Thématique : Image

MAJA

MAJA (MACCS ATCOR Joint Algorithm)

The MAJA (MACCS ATCOR Joint Algorithm, pronounced "maya") chain is a cloud detection and atmospheric correction chain. It is adapted to the processing of time series of high resolution images acquired under constant or quasi constant viewing angles. It can be used to process data from LANDSAT and SENTINEL-2.  It is based on the MACCS chain developed by CNES and CESBIO since 2008, and gradually incorporates since 2016 methods from the ATCOR chain of DLR. It is now the subject of a collaboration between CNES, DLR and CESBIO, and is financially supported by ESA.

Its particularity is to use the multi-temporal information contained in the images to detect clouds and their shadows, to estimate the atmospheric content in aerosols and water vapor, and to correct the atmospheric effects (taking into account the environmental and relief effects).

For more details on the methods associated with this chain: http://www.cesbio.ups-tlse.fr/multitemp/?p=6050

How to run MAJA ?

To process time series, an example of MAJA chain is available here: https://github.com/CNES/Start-MAJA

Data format ?

Maja 3.2.2 is provided in two versions for the Sentinel-2 platform:

TM :A version capable of outputting 2A products in Theia format (described here) from level 1C products available at ESA or on PEPS.
noTM : A version that outputs 2A products in "native" format (described here), again from Level 1C products available at ESA or on PEPS. This version is provided to maintain continuity for MAJA 1.0 users, but this will be discontinued in the near future. New users are advised to use the TM version.
Maja 1.0 only handles the native 2A format.

Inputs:

The delivered version of MAJA is compatible with L1 products in the following formats (documents provided in the package):

Sentinel-2: (PSD) S2-PDGS-TAS-DI-PSD
Landsat 8: Landsat8_Level1_Data_Format_Control_Book_LSDS-809
Landsat 5-7: CNES internal format described in PSC-IF-411-0081-CNES_E1R4
Venus: Format description
Installation :

Unlike version 1.0, the libraries needed to run MAJA 3.2.2+ are precompiled and provided in the executable.

To install MAJA, after unzipping the downloaded file, you have to run the command

>>> bash MAJA-3.2.2_TM.run --target /path/to/install

The tool will give you the following answer:
          Verifying archive integrity...  100% All good.

          Uncompressing MAJA 3.2.2 100%  

          You can now run
          /home/hagolle/PROG/MAJA_Distrib/Maja-3.2.2-TM/bin/maja

          List of spectral camera enabled:

          LANDSAT8 
More documentation can be found in
          /home/hagolle/PROG/MAJA_Distrib/Maja-3.2.2-TM/README
Sentinel2 does not appear in the list (this is a bug) but it is present.

Usage

We recommend to use the following script to process your time series: https://github.com/CNES/Start-MAJA

The script documentation also indicates how to get the right execution parameters and prepare all the necessary inputs.

MAYA

MAJA

Thematic

IMAGE

Licence type

Open source

Programming language

C++; Python; Cmake

Operating system

Maja 3.2.2: Linux RedHat 6+, CentOS 6+, Ubuntu 12+

Click to choose software version
contact image

A question regarding MAJA software ?

Don't hesitate to contact us on maja-support@cnes.fr address ! We are here to help you.

Contact

Stay connected

Receive the latest software news of MAJA