An openEO back-end driver for geospatial data science using the Pangeo software stack

Open Source code from Geosoft II seminar at ifgi, winter semester 2019/20

Organisation page

About the course

During the winter semester 2020/21 the bachelor students at the Institute for Geoinformatics (ifgi) were challenged to develop an openEO back-end driver using latest Python-based numerical processing libraries. The openEO project developed an API for exposing the power of big Earth observation cloud infrastructures in a standardised way based on the data cube paradigm. The openEO API allows users to easily switch between service and infrastructure providers and use the most suitable or cost-effective solution. In the "story" of the seminar, a startup wants to provide a new implementation of the openEO API based on the Pangeo software stack (xarray, Dask) to its customers and the students are hired to do just that. The group achieved quite impressive results, including a complex CI test setup, nicely demonstrated in a number of Jupyter Notebooks.

You can download the full project requirements document for details.


The course was taught by Christian Knoth, Daniel NĂ¼st and Edzer Pebesma.

LSF Nr.: 142952 —  Learnweb


Demos

https://github.com/Geosoft2/Demos/

API project with ready-to-use docker-compose configuration

https://github.com/Geosoft2/GeoSoftII_Projekt

Job JSON screenshot

Microservices for processes and management

https://github.com/Geosoft2/GeoSoftII_SST_Process
https://github.com/Geosoft2/GeoSoftII_DataServer
https://github.com/Geosoft2/GeoSoftII_JobManagement
https://github.com/Geosoft2/GeoSoftII_Frontend
https://github.com/Geosoft2/GeoSoftII_NDVI_Process

SST

Test repository using GitHub actions

https://github.com/Geosoft2/TestRepo

Screenshot