Rilevamento di rocce da immagini satellitari

The modern digital computer has been fundamental to the space exploration program. Before computers, exploring space was quite difficult. Almost everything in space exploration is done with computers. Computers have also led to many major break- throughs in space research. From designing spaceships to space photography, almost everything is done with computers. There are thousand of human devices such as telescopes, satellites, landers and rovers that are gathering data from Earth, space and even other planets. Each of these device produces a large amount of data. The main problem is that the data need to be processed and although computers help speed up this process, there is still a need for a human to process that data, this requires a lot of time to be done so only a small part of the data is actually used and processed. In recent years with the advent of Artificial Intelligence, in particular Machine Learning, scientists have begun to use this technology to improve space exploration. AI has proven its great potential and is a game-changer in space exploration such as charting unmarked galaxies, stars, black holes, and studying cosmic events, as well as communication, autonomous starcraft navigation, monitoring and system control. The most recent use case of AI is found in the endeavors to create AI-powered, empathetic robotic assistants to help astronomers in their long space travel by understanding and predicting the crew’s needs and comprehending astronauts’ emotions. By using AI we can use the small amount of data processed by scientists to train algorithms which can process enormous amount of data faster than a person. One of the fields of Machine Learning used for space exploration is Computer Vision which allows computers and systems to derive meaningful information from digital images, videos, and other visual inputs - and take action or report on that information. If AI allows computers to think, computer vision allows them to see, observe and understand. The exact Computer Vision technique that I studied is called Object Detection which allows to identify and locate objects in images and videos. The aim is to find an Object Detection Algorithm able to learn to detect complex objects such as rocks from satellite images, in our case the images were taken from a satellite on Enceladus, the sixth-largest moon of Saturn. But, why is it useful to detect the position of rocks in planets or moons? For a Lander, knowing the position of rocks helps indicate the best place where to land, while for a rover it can indicate the best route to proceed along. To a geologist, rocks provide clues for the paleohistory of the planet. Depending on the type and distribution of rocks (e.g. igneous, sedimentary or metamorphic) found, a scientist can deduce what the area was like the time the rocks were being formed and deposited.