About 109,000 previously unrecognised impact craters on the moon using machine learning techniques have been identified by a team of researchers.
The study, led by Jilin University researchers, was published in the journal Nature Communications.
Effect craters are the most prominent feature on the lunar surface and cover most of the surface of the moon.
Standard automated detection methods typically make it impossible to identify unusual and badly damaged impact craters that may have formed in the early stages.
In order to accurately classify craters and predict their age, researchers used a transfer learning approach and trained a deep neural network with data from previously identified craters, reports Xinhua News Agency.
By integrating the data obtained by the Chinese Chang’e-1 and Chang’e-2 lunar probes, the researchers found 109,956 new impact craters. They also measured the age of 18,996 recently detected craters greater than 8 km in diameter.
In the meantime, researchers have developed a new database of lunar impact craters for the mid-and low-latitude regions of the moon.
Yang Chen of the University of Jilin, who is one of the experts, said that the database of lunar craters is of great importance to scientific science on the moon.
The embraced technique can be extended to endorse crater studies, producing credible suggestions for planetary science,” said Yang.