Customer: Centre National d’Études Spatiales (CNES)

Programme: Generic

Supply Chain: CNES > CS Group SPACE

Context

CS Group responsabilities for Orfeo Toolbox are as follows:

• Library maintenance and upgrades
• Community Support
• Animation of communities of users and developers
• Maintenance of the platform for online services

The features are as follows:

• Segmentation: applicable on a large scale (mean shift, watersheds, etc.) vectorization, comparison Machine Learning: Training of classification and regression models. Application on image and vector data. Assessment Extraction of primitives: Contour detection, calculation of radiometric indices, texture extraction, calculation of global and local statistics, mathematical morphology Image Geometry: Ortho-Rectification, Model Sharpening, Grid Resampling, PanSharpening, Disparity Calculation, Image Registration Calibration, Hyperspectral…

Project implementation

The project objectives are as follows:

• OTB: Free image processing toolbox for remote sensing, consisting of a C++ library and a set of applications accessible from different Wrappers: Python 3, command line, QGIS plugin, GUI.

The processes for carrying out the project are:

• Iterative development
• Continuous development and integration (coding/unit testing)
• Automatic Test Plan Generator with test execution and insertion of results in a CDash instance

Technical characteristics

The solution key points are as follows:

• C++ software based on OTB libraries
• Python
• CMake / CTest / CDash
• Managing multiple plugins using generic factories

The main technologies used in this project are:

Domain	Technology(ies)
Operating System(s)	Linux, Windows
Programming language(s)	C++ 14, Python
Production software (IDE, DEVOPS etc.)	CMake, SWIG, Gitlab, Gitlab-CI, Sphinx, Discourse, Wordpress, Doxygen
Main COTS library(ies)	ITK, GDAL, Boost, 6S, MuParser, Shark ML