With the cooperation of the University of Bern – Artificial Intelligence is to track down new exoplanets
Artificial intelligence is supposed to track down new exoplanets
An algorithm developed by researchers from the Universities of Bern and Geneva is supposed to recognize every pixel of a constellation so that new exoplanets can be discovered.
An artificial intelligence (AI) based image recognition method of the planets outside our solar system on which new approaches remain invisible. Astronomers from Bern and Geneva report on this in the specialist magazine “Astronomy and Astrophysics”.
Most of the more than 4800 exoplanets discovered so far by astronomers have been identified using the so-called transit method. “Transit” is what experts call the phenomenon when, from the observer’s point of view, a planet passes in front of its star and obscures it. Only: interactions between planets can change the periodicity of the transit and thus make a discovery impossible. This is because the mother star is not eclipsed in a periodic rhythm.
A team from the University of Bern and Geneva in cooperation with the company Disaitek are working on a solution to this problem: They taught a machine to predict the effects of this planetary generation.
Characterization of the exoplanets
By supplying the AI algorithm with countless training images, it learns to recognize for each pixel of a constellation which object it represents – for example, an eclipse by a planet.
“When we first used the method, we discovered two exoplanets – Kepler-1705b and Kepler-1705c – that were completely overlooked by earlier techniques,” said Adrien Leleu of the University of Geneva, according to a joint communication from the universities on Thursday. The method can be used to estimate the radius of the planet and provides information about its mass, density and composition.
The new technology could prove useful not only for astronomy, but also for observing the earth and its environment. So Disaitek is now using the AI to track down illegal landfills in high-resolution satellite images.
Found a mistake?Report now.