Artificial Intelligence Machine Learning Rille Moon Landing Exploration
Scientists Siyuan Chen, Xin Gao and Shuyu Sun as well as researchers from The Chinese University of Hong Kong, are using applied machine learning and artificial intelligence to automate the identification of potential landing as well as exploration sites on the Moon. Their goal is to look for lunar features such as craters and rilles, which are allegedly hot spots for energy resources like uranium and helium-3—a for nuclear fusion. Read more for a short video and additional information.



Both uranium and helium-3—a have been detected in moon craters and could be very useful resources for replenishing spacecraft fuel. One issue they encountered was that there was no labeled dataset for rilles, so the scientists had to construct their own training dataset with annotations for both craters as well as rilles. They used an approach called transfer learning to pre-train their rille model on a surface crack dataset with some fine tuning using actual rille masks. Next, they developed a computational approach to identify craters and rilles simultaneously.

Celestron - NexStar 8SE Telescope - Computerized Telescope for Beginners and Advanced Users - Fully-Automated GoTo Mount - SkyAlign Technology - 40,000+ Celestial Objects - 8-Inch Primary Mirror
1,781 Reviews
Celestron - NexStar 8SE Telescope - Computerized Telescope for Beginners and Advanced Users - Fully-Automated GoTo Mount - SkyAlign Technology - 40,000+ Celestial Objects - 8-Inch Primary Mirror
  • Nexstar computerized telescope: The NexStar 8SE Computerized Telescope features Celestron’s iconic orange tube design with updated technology and...
  • 8-Inch aperture: The 8-inch primary mirror in this Schmidt-Cassegrain telescope for adults and kids to be used together packs enough light-gathering...
  • Fully-automated go to mount: Featuring a database of more than 40,000 celestial objects, the go to mount built into our telescopes for astronomy...

This is a pixel-to-pixel problem for which we need to accurately mask the craters and rilles in a lunar image. We solved this problem by constructing a deep learning framework called high-resolution-moon-net, which has two independent networks that share the same network architecture to identify craters and rilles simultaneously,” said Siyuan Chen.

Write A Comment