Mars visual simulation for ExoMars navigation algorithm validation

The ExoMars rover will make heavy use of machine vision algorithms for autonomous navigation. To support the validation of these algorithms, a visual simulation capable of generating realistic simulated images of a Mars-like virtual environment has been developed. This is an enhancement of the University of Dundee’s PANGU planet surface simulation tool. The simulation is based on data from the planetary science literature and validated using images from the Mars Exploration Rovers. It includes models of rocks, surface reflectance, atmospheric dust, a rover vehicle and dynamically computed shadows. It provides a straightforward way for autonomy researchers to generate high-quality images for testing, and has potential to be used as a more general rover simulation visualisation tool