Related Experiment Video
Updated: Nov 24, 2025
04:48
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
Published on: November 30, 2022
3.1K
Lunar impact crater identification and age estimation with Chang'E data by deep and transfer learning
Chen Yang1,2, Haishi Zhao3, Lorenzo Bruzzone4
1College of Earth Sciences, Jilin University, 130061, Changchun, China. yangc616@jlu.edu.cn.
Nature Communications
|December 23, 2020
Summary
Scientists automatically detected over 100,000 new lunar craters using deep neural networks and Chang
Area of Science:
- Planetary Science
- Geology
- Remote Sensing
Background:
- Impact craters are crucial for understanding Solar System history.
- Existing lunar crater databases are limited in size and scope.
- Automatic detection and age estimation are needed to advance lunar science.
Purpose of the Study:
- To develop an automated method for detecting and estimating the ages of lunar impact craters.
- To significantly expand the number of known lunar craters and their age data.
- To create a comprehensive lunar crater database for scientific research.
Main Methods:
- Utilized transfer learning with deep neural networks.
- Processed Chang'E mission data combined with stratigraphic information.
- Progressively identified new craters from an initial dataset.
Main Results:
- Identified 109,956 new lunar craters, a more than twelvefold increase.
- Estimated formation ages for 18,996 craters larger than 8 km.
- Created a new lunar crater database for mid- and low-latitude regions.
Conclusions:
- Automated crater detection using deep learning significantly enhances lunar surface mapping.
- The new database provides valuable data for planetary science and Solar System studies.
- This methodology can be applied to other celestial bodies for geological analysis.
