Communications of the ACM
Computer Searches Telescope Data for Evidence of Distant Planets
The U.S. National Aeronautics and Space Administration (NASA) has established a crowdsourcing project in which volunteers search telescopic images for evidence of debris disks around stars, which are good indicators of exoplanets.
Credit: NASA/JPL-Caltech
Massachusetts Institute of Technology (MIT) researchers have trained a machine learning system to search for debris disks around stars using telescopic data.
The system was found to corroborate with human identifications of debris disks 97% of the time, while the team also trained it to rate debris disks it found based on their likelihood of containing detectable exoplanets.
The researchers say the system identified 367 previously unexamined celestial objects as promising candidates.
MIT's Tam Nguyen first carved up the data into small chunks, then applied standard signal-processing techniques to sift out artifacts. Afterwards, she identified chunks with light sources at their centers, and used existing image-segmentation algorithms to excise any additional sources of light. Nguyen used basic principles of physics to filter the data further, studying variation in the intensity of the light emitted across four different frequency bands, and using standard metrics to evaluate the position, symmetry, and scale of the light sources.
From MIT News
View Full Article
Abstracts Copyright © 2018 Information Inc., Bethesda, Maryland, USA
No entries found