Image processing for Earth observation

Course Summary and Goals

This course aims at getting through all the major steps related to image processing for remote sensing of the Earth, from sensors to machine learning, including:

1. Basic concepts of remote sensing and digital imaging of the Earth
2. Platforms, sensors & image characteristics
3. Information extraction
4. Image segmentation and classification with machine learning
5. Deep learning methods

Organisation and Timing

Each week, you will find all the course material (slides, exercises, references, ...) on moodle and you will have to:

• study reference material (slides, links to websites, papers or videos to watch). 
• (on friday morning, first 1-2 hours) attend the lecture sessions 
• (on friday morning, last 1-2 hours) attend the exercises, where we can discuss about the exercises, the project and deepen some of the lecture aspects.
• All the 4 hours are both performed on campus. We will provide recordings from the 2021 edition later in the semester.  
• Recordings of the lectures will be uploaded and the link placed in the corresponding week section.

How to stay tuned

The 'News forum' is used by the teaching team to make announcements to the class.  Only the teaching team will post in announcements.

  News forum

We use a Slack channel for quick communication. 

• Join the Slack channel of the course by using this link: https://join.slack.com/t/epfl-6bz3079/shared_invite/zt-2qb64ot6q-E71lLdjrRMBL4oY3Q14BvQ

Assessment 

The objectives of the course will be assessed as follows:

• A mid term is scheduled on November 29, 2024, in the classroom (50% of the grade).

• At the end of the course, group of students will submit their project composed of a final report and appendix: code, etc. (50% of the grade). The project reports deadline will be further communicated, but as a reference the deadline will be set before the winter exams session starts