teach_2024_msc_envscience_remote_sensing_and_geoinformatics

Teaching material for the MSc Module "Remote Sensing and Geoinformatics"

News

Wednesday

• Maximilian Fabi will show you the orthoimage processing. He will pick you up at CIP3 at 09:15 (usual place and time).
• Finish the leftovers of the citizen science assignments or relax :-)

Thursday

• you should read https://www.sciencedirect.com/science/article/pii/S2667393223000054
  • (no need to understand every piece and term but you should have a rough overview)
• checkout the current state of the database we are currenlty setting up at https://www.deadtrees.earth
• Delindate orthoimagery; the idea was 2-5 Orthoimages per person ... depending on how much you want to support us :-)
  • A HowTo on the delineation process and the data can be found here.

Friday

• Get a glimpse on why deep learning-based pattern recognition is so powerful in the remote sensing domain. You may checkout our publication in that context, e.g,:
  • https://www.sciencedirect.com/science/article/pii/S0924271620303488
  • https://www.sciencedirect.com/science/article/pii/S0924271620302938
  • (no need to understand every piece and term but you should have a rough overview)
• Compare your delineated orthoimages to our artifical intelligence (pattern recognition using Convolutional Neural Networks)
• This might only work in Colab, as it requires GPU resources. Find the script here
• Brainstorm on what you would like to do as a individual group project next week (2 persons per group).

Small list of advices or this course:

• dare to ask many questions
• dare to give feedback
  • in the course, to teja.kattenborn@geosense.uni-freiburg or also here in Github unter "Issues" or "Disscussions)
  • For instance, feedback on improvements for the course, errors in the code, analysis of interest, etc.
• Be patient with yourself (expect both flat and steep learning curves)
• Help eachother :-)

Access to the data

• All data can be found here.
• In total the data amounts to 18 GB. Let me know if you do not have sufficient storage ressources for the duration of the course.

Where to perform the analysis?

• Your own computer
• CIP-Pool computers
• Google Colab: https://colab.research.google.com/
• Colab-JuPyteR shortcuts: https://colab.research.google.com/drive/13IO3-gfyS9mSPuzAo6-wsYBUOVpxb_va?usp=sharing

install packages

• pip install rioxarray (in colab)
• alternatives conda or mamba (conda install -c conda-forge)
• list of packages that should be installed for the course: rioxarray rioxarray matplotlib numpy xarray glob2 pandas geopandas rasterstats rasterio