Two-Step super-resolution method to enhance orbital imagery of celestial bodies using ground truth from rovers and drones.

Exploration of other planetary bodies poses multiple challenges. Notably, how can we obtain the most information about a planet or a moon while avoiding risky and costly crewed or robotic missions to the surface? Satellite imagery can partly solve this problematic by providing a relatively cheap alternative to cover massive areas of a planet. Sounds great, what’s the catch? Well, these satellite images have a limited resolution and therefore, limited information can be extracted from them. That being said, recent advances in artificial intelligence could offer us a way to enhance the resolution of satellite imagery. This projects consequently aims to test various artificial intelligence algorithms to help enhance the resolution of Martian orbital imagery. In term, this method could help the Canadian Space Agency and private partners such as Mission Control Space Services select region of particular interest for further investigation.

Faculty Supervisor:

Myriam Lemelin

Student:

Partner:

Mission Control

Discipline:

Earth science

Sector:

Manufacturing; Professional, scientific and technical services

University:

Université de Sherbrooke

Program: