학술논문

Lunar Shadowlands.
Document Type
Article
Author
Source
Scientific American. Feb2022, Vol. 326 Issue 2, p16-16. 2/3p. 1 Color Photograph.
Subject
*DEEP learning
*ALGORITHMS
*LUNAR geography
*SOLAR system
PHOTOGRAPHS of the moon from space
Language
ISSN
0036-8733
Abstract
The article focuses on a deep-learning algorithm developed by researchers to photograph the permanently shadowed regions (PSR) near the poles of the moon in a study published in the periodical "Nature Communications." PSR contain ice reservoirs that could provide data on the early solar system. Planetary scientist Valentin Bickel cites the ability of the algorithm to produce images that identify geologic features. It mentions the use of camera temperature and position to train the algorithm.