With over 60 known moons, Saturn and its rings has captivated scientists for hundreds of years. By applying image recognition and machine learning algorithms, scientists are hoping to accelerate their understanding of the sixth planet from the Sun.
Using actual images from the Cassini spacecraft, the goal is to create an algorithm for scientific advancement to better understand ring phenomena, ring structure, and potentially find new moons that are not otherwise detectable by a computer due to false-positives.
We segregated this project into 2 distinct challenges to arrive at a valuable solution. The 1st challenge established a solid approach to solving the problem, allowing the competitors in the 2nd challenge to focus solely on the accuracy of their algorithmic solution.
The winning algorithm was run against ~30,000 of the highest resolution Cassini images. The tool identified 81% of those previously identified propellers while helping to identify never before discovered propellers in the rings of Saturn. To date, at least 4 detections of previously unseen propellers have been identified by the new algorithmic solution.
The new data solution will continue to accelerate the discovery within the rings of Saturn and far beyond.