Automatic Segmentation of Radar Data from the Chang’E-4 Mission Using Unsupervised Machine Learning - A Data-Driven Interpretation Approach

Aug 5, 2026

Upload by Ivy

Chang’E-4 mission ground-penetrating radar (GPR) interpretation remains challenging due to the Moon’s subsurface complexity and heterogeneity, producing difficult-to-read radagrams with low signal-to-clutter ratios. Clutter obscures potential targets and makes traditional interpretation highly laborious and subjective. The proposed framework applies unsupervised machine learning by computing local statistical attributes over sliding windows and clustering them into formation groups. Fully automatic segmentation highlights boundaries between formations and reveals hidden structures.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages17

Reading Time43 min

AccessFree to download

Abstract

Keywords

Introduction

Edinburgh Research Explorer Automatic Segmentation of Radar Data from the Chang'E-4 Mission Using Unsupervised Machine Learning: A Data-Driven Interpretation Approach Citation for published version: Giannakis, I, Mcdonald, C, Feng, J, Zhou, F, Su, Y, Martin-Torres, J, Zorzano, MP, Warren, C, Giannopoulos, A & Leontides, G 2024, 'Automatic Segmentation of Radar Data from the Chang'E-4 Mission Using Unsupervised Machine Learning: A Data-Driven Interpretation Approach', Icarus, vol. 417, 116108. [https://doi.org/10.1016/j.icarus.2024.116108](https://doi.org/10.1016/j.icarus.2024.116108) Digital Object Identifier (DOI): 10.1016/j.icarus.2024.116108 Link: Link to publication record in Edinburgh Research Explorer Document Version: Peer reviewed version Published In: Icarus General rights Copyright for the publications made accessible via the Edinburgh Research Explorer is retained by the author(s) and / or other copyright owners and it is a condition of accessing these publications that users recognise and abide by the legal requirements associated with these rights. Take down policy The University of Edinburgh has made every reasonable effort to ensure that Edinburgh Research Explorer content complies with UK legislation. If you believe that the public display of this file breaches copyright please [contact openaccess@ed.ac.uk](contact openaccess@ed.ac.uk) providing details, and we will remove access to the work immediately and investigate your claim. Download date: 04. Jul. 2024 AUTOMATIC SEGMENTATION OF RADAR DATA FROM THE CHANG’E-4 MISSION USING UNSUPERVISED MACHINE LEARNING: A DATA-DRIVEN INTERPRETATION APPROACH Iraklis Giannakis, University of Aberdeen, School of Geosciences Aberdeen, United Kingdom [iraklis.giannakis@abdn.ac.uk](iraklis.giannakis@abdn.ac.uk) Ciaran McDonald, RSK Geophysics Hemel Hempstead, United Kingdom [ciaran_mcdonald@outlook.com](ciaran_mcdonald@outlook.com) Jianqing Feng Planetary Science Institute Denver, USA [jfeng@psi.edu](jfeng@psi.edu) Feng Zhou China University of Geosciences (Wuhan) Wuhan, China [zhoufeng@cug.edu.cn](zhoufeng@cug.edu.cn) Yan Su Chinese Academy of Sciences Beijing, China [suyan@nao.cas.cn](suyan@nao.cas.cn) Javier Martin-Torres Maria-Paz Zorzano University of Aberdeen, School of Geosciences Centro de Astrobiologia, CSIC-INTA, Torrejon de Ardoz Aberdeen, United Kingdom Madrid, Spain [javier.martin-torres@abdn.ac.uk](javier.martin-torres@abdn.ac.uk) [zorzanomm@cab.inta-csic.es](zorzanomm@cab.inta-csic.es) Craig Warren Antonios Giannopoulos Northumbria University The University of Edinburgh Northumbria, United Kingdom Edinburgh, United Kingdom [craig.warren@northumbria.ac.uk](craig.warren@northumbria.ac.uk) [a.giannopoulos@ed.ac.uk](a.giannopoulos@ed.ac.uk) Georgios Leontidis Interdisciplinary Centre for Data & AI, University of Aberdeen Aberdeen, United Kingdom [georgios.leontidis@abdn.ac.uk](georgios.leontidis@abdn.ac.uk) ABSTRACT 1 Chang’E-3, E-4 and E-5 were the first planetary missions with in-situ ground-penetrating radar (GPR) 2 in their scientific payloads. Apart from the Chang’E missions, in-situ GPR has also been used in the 3 Martian missions Tianwen-1 and Perseverance, establishing GPR as a mainstream and pivotal tool in 4 the new era of planetary exploration. Despite its widespread use in planetary science, interpreting 5 GPR data from the Chang’E-4 mission is still challenging, with varying and ambiguous interpretations 6 offered by different researchers. This is primarily due to the complexity and heterogeneity of the 7 Lunar subsurface, which results to a difficult to interpret radagram with low signal-to-clutter ratio. 8 Clutter masks potential targets and makes interpretation a laborious and highly subjective process. To 9 tackle this, we interpret the GPR data from Chang’E-4 mission using a novel processing paradigm 10 based on un-supervised machine learning. A

FAQs

Why is interpreting Chang’E-4 GPR data difficult?

The lunar subsurface is complex and heterogeneous, leading to radagrams with low signal-to-clutter ratio. Clutter masks targets and makes interpretation subjective and time-consuming.

How does the proposed method perform segmentation?

It computes statistical attributes locally for sliding windows on the radagram, then clusters these features using unsupervised machine learning. The output highlights boundaries between different formations.

What benefits does the framework provide over typical processing approaches?

It is fully automatic and objective, improving processing reliability. It also reveals hidden structures that may be unseen using conventional approaches.

Aug 5, 2026

Upload by Ivy