Computer Vision in Space

About

Computer vision in space involves the application of image processing and artificial intelligence techniques to analyse visual data obtained in space environments. This field encompasses the development and implementation of algorithms for tasks such as object recognition, tracking, and autonomous navigation in space missions. It plays a crucial role in interpreting satellite imagery, aiding planetary exploration through robotic missions, and enhancing the capabilities of spacecraft and rovers. The goal is to enable machines to "see" and interpret visual information, contributing to the success and efficiency of various space-related endeavours.

The course on Computer Vision in Space aims to provide a comprehensive understanding of various aspects related to the application of visual methods towards space exploration. The course offers insights into the algorithms and methods employed in space applications. During the two-week programme, participants will delve into the topic’s foundational principles, deployment methods and approaches to operate onboard spacecrafts or rovers. Exercises aligned with the lectures will provide the required hands on experience, which will be tied into a short summarizing project work at the end of the summer school, where the participants can apply the various aspects learned during the last days.

An exciting and unique aspect of the course is the opportunity for participants to test their own implementations on the KuupKulgur, the Estonian lunar rover platform. This practical exposure will provide a rare chance for students to witness the translation of theoretical concepts into real-world scenarios. Throughout the week, several experts in the field will present inspirational talks. Additionally, there will be a guided visit to the Tartu Observatory Space Missions Simulation Center.

Course leader

Ric Dengel, University of Tartu

Target group

MA/PhD and advanced BA

Course aim

1. Develop a strategy to solve a specific computer vision problem based on known algorithms and deployment on an applicable processing platform.
2. Understand and analyse different processing architectures for space computer vision applications.
3. Demonstrate the ability to deploy different algorithms on a space processing platform.
4. Demonstrate and apply knowledge of computer vision techniques towards space applications.
5. Understand the trade-off between different geometrics and non-geometric approaches in computer vision methods.

Fee info