Unsupervised TomoSAR Image Reconstruction Through Virtual Multiple Measurement Explorations

Tomographic synthetic aperture radar (TomoSAR) reconstruction produces 3D imaging of scenes from measurements and has recently been combined with data-driven deep learning techniques. While most existing methods rely on supervised learning with simulation-based and paired training samples, we focus...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Proceedings of the IEEE National Radar Conference (1996) S. 321 - 326
Hauptverfasser: Liu, Liang, Zeng, Tianjiao, Wang, Mou, Shi, Jun, Wei, Shunjun, Zhang, Xiaoling, Zhan, Xu
Format: Tagungsbericht
Sprache:Englisch
Veröffentlicht: IEEE 04.10.2025
Schlagworte:
ISSN:2375-5318
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Abstract Tomographic synthetic aperture radar (TomoSAR) reconstruction produces 3D imaging of scenes from measurements and has recently been combined with data-driven deep learning techniques. While most existing methods rely on supervised learning with simulation-based and paired training samples, we focus on unsupervised learning that operates without such samples. This approach promises more robust, flexible, and generalized reconstruction capabilities while leveraging deep learning to enhance reconstruction quality. We present a preliminary exploration using virtual multiple measurements derived from raw measurements. We leverage the recent finding that imaging areas possess both non-local and local similarities, resulting in highly correlated measurements. This property enables us to obtain additional virtual measurements beyond the raw data. By treating these measurements as multiple independent noisy samples of the same underlying signal, we propose a self-supervised robust reconstruction loss that incorporates an integrated physics forward measurement model. Our approach feeds measurement pairs to both the input and output of the reconstruction neural network, enabling it to explore the underlying signal while suppressing noise. We also adopt an encoderdecoder network architecture that implicitly incorporates scene priors through its inductive bias. By combining the physics model prior and scene prior while remaining robust to noise, we achieve high-fidelity reconstruction results. Experiments on both simulated and measured data demonstrate our method's effectiveness, achieving reconstruction performance comparable to supervised approaches while showing better generalization under varying noise conditions.
AbstractList Tomographic synthetic aperture radar (TomoSAR) reconstruction produces 3D imaging of scenes from measurements and has recently been combined with data-driven deep learning techniques. While most existing methods rely on supervised learning with simulation-based and paired training samples, we focus on unsupervised learning that operates without such samples. This approach promises more robust, flexible, and generalized reconstruction capabilities while leveraging deep learning to enhance reconstruction quality. We present a preliminary exploration using virtual multiple measurements derived from raw measurements. We leverage the recent finding that imaging areas possess both non-local and local similarities, resulting in highly correlated measurements. This property enables us to obtain additional virtual measurements beyond the raw data. By treating these measurements as multiple independent noisy samples of the same underlying signal, we propose a self-supervised robust reconstruction loss that incorporates an integrated physics forward measurement model. Our approach feeds measurement pairs to both the input and output of the reconstruction neural network, enabling it to explore the underlying signal while suppressing noise. We also adopt an encoderdecoder network architecture that implicitly incorporates scene priors through its inductive bias. By combining the physics model prior and scene prior while remaining robust to noise, we achieve high-fidelity reconstruction results. Experiments on both simulated and measured data demonstrate our method's effectiveness, achieving reconstruction performance comparable to supervised approaches while showing better generalization under varying noise conditions.
Author Zeng, Tianjiao
Wei, Shunjun
Zhang, Xiaoling
Zhan, Xu
Liu, Liang
Shi, Jun
Wang, Mou
Author_xml – sequence: 1
  givenname: Liang
  surname: Liu
  fullname: Liu, Liang
  organization: University of Electronic Science and Technology of China,Chengdu,China,611731
– sequence: 2
  givenname: Tianjiao
  surname: Zeng
  fullname: Zeng, Tianjiao
  email: tzeng@uestc.edu.cn
  organization: University of Electronic Science and Technology of China,Chengdu,China,611731
– sequence: 3
  givenname: Mou
  surname: Wang
  fullname: Wang, Mou
  organization: University of Electronic Science and Technology of China,Chengdu,China,611731
– sequence: 4
  givenname: Jun
  surname: Shi
  fullname: Shi, Jun
  organization: University of Electronic Science and Technology of China,Chengdu,China,611731
– sequence: 5
  givenname: Shunjun
  surname: Wei
  fullname: Wei, Shunjun
  organization: University of Electronic Science and Technology of China,Chengdu,China,611731
– sequence: 6
  givenname: Xiaoling
  surname: Zhang
  fullname: Zhang, Xiaoling
  organization: University of Electronic Science and Technology of China,Chengdu,China,611731
– sequence: 7
  givenname: Xu
  surname: Zhan
  fullname: Zhan, Xu
  email: zhanxu@std.uestc.edu.cn
  organization: University of Electronic Science and Technology of China,Chengdu,China,611731
BookMark eNo1UF9LwzAcjKLgNvcNfMiTb5350yzJ4xhTBxtCrb74MNLm163SJiVpRb-9FfXlDu64g7spunDeAUK3lCwoJfouM9aEtXcVE0ITJReMMDFajAjKyRmaa6kV51Sk6YjnaMK4FIngVF2haYzvhAg-ShP09uLi0EH4qCNYnPvWP68yvG3NEXAGpXexD0PZ197h_BT8cDzh1zr0g2nwfmj6umsA78HEIUALrsebz67xwfwE4jW6rEwTYf7HM5Tfb_L1Y7J7etiuV7uk1rxPJCspVAWzSi-tUSUpwBZMFqIaNyphlKJWppJWAigoW0peiFQvSylSI1Kr-Azd_NbWAHDoQt2a8HX4_4J_Az9iWus
ContentType Conference Proceeding
DBID 6IE
6IH
CBEJK
RIE
RIO
DOI 10.1109/RadarConf2559087.2025.11205130
DatabaseName IEEE Electronic Library (IEL) Conference Proceedings
IEEE Proceedings Order Plan (POP) 1998-present by volume
IEEE Xplore All Conference Proceedings
IEEE Electronic Library (IEL)
IEEE Proceedings Order Plans (POP) 1998-present
DatabaseTitleList
Database_xml – sequence: 1
  dbid: RIE
  name: IEEE Electronic Library (IEL)
  url: https://ieeexplore.ieee.org/
  sourceTypes: Publisher
DeliveryMethod fulltext_linktorsrc
Discipline Engineering
Physics
EISBN 9798331544331
EISSN 2375-5318
EndPage 326
ExternalDocumentID 11205130
Genre orig-research
GrantInformation_xml – fundername: Aeronautical Science Foundation of China
  grantid: 1A2024Z071080005
  funderid: 10.13039/501100004750
– fundername: National Natural Science Foundation of China
  grantid: 62471113,62305049,62371104
  funderid: 10.13039/501100001809
GroupedDBID 6IE
6IF
6IH
6IK
6IL
6IM
6IN
AAJGR
AAWTH
ABLEC
ADZIZ
ALMA_UNASSIGNED_HOLDINGS
BEFXN
BFFAM
BGNUA
BKEBE
BPEOZ
CBEJK
CHZPO
IEGSK
IJVOP
IPLJI
OCL
RIE
RIL
RIO
ID FETCH-LOGICAL-i93t-72c1efb2d896da8c0bedb27b5fada85a881d7471f5e1e8dc73b5496c754a54d83
IEDL.DBID RIE
IngestDate Wed Nov 05 07:15:55 EST 2025
IsPeerReviewed false
IsScholarly true
Language English
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-i93t-72c1efb2d896da8c0bedb27b5fada85a881d7471f5e1e8dc73b5496c754a54d83
PageCount 6
ParticipantIDs ieee_primary_11205130
PublicationCentury 2000
PublicationDate 2025-Oct.-4
PublicationDateYYYYMMDD 2025-10-04
PublicationDate_xml – month: 10
  year: 2025
  text: 2025-Oct.-4
  day: 04
PublicationDecade 2020
PublicationTitle Proceedings of the IEEE National Radar Conference (1996)
PublicationTitleAbbrev RADARCONF25
PublicationYear 2025
Publisher IEEE
Publisher_xml – name: IEEE
SSID ssj0053237
Score 2.3053691
Snippet Tomographic synthetic aperture radar (TomoSAR) reconstruction produces 3D imaging of scenes from measurements and has recently been combined with data-driven...
SourceID ieee
SourceType Publisher
StartPage 321
SubjectTerms data-driven
Deep learning
Image reconstruction
Loss measurement
neural network inductive bias
Noise
Noise measurement
Physics
Redundancy
Three-dimensional displays
tomoSAR
Training
Unsupervised learning
Title Unsupervised TomoSAR Image Reconstruction Through Virtual Multiple Measurement Explorations
URI https://ieeexplore.ieee.org/document/11205130
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV09T8MwED1BBQgWoLTiWx4QW9omjmNnRIgKhlZVCagSQ-XYF6lDk6pp-f3YTkphYGCLrChyztHdvct7dwB3BhLETDLh6VgqL1SSeQJ932PCxA5NUUSpcsMm-HAoJpN4VIvVnRYGER35DDv20v3L14Va21JZ1-QG5huiBqHvch5VYq2N22U0oPwA7usmmt2xNDjcquZsztwT3EDBgHU2T_g1S8WFkv7xPzdxAq2tKI-MvsPNKexg3oSjH_0Em7Dv-JyqPIOPt7xcL6wfKFGTpJgXrw9j8jI33oNYxLntG0uSalIPeZ8trZaEDGqKIRlsy4ekoupV1b0WJP2n5PHZq-coeLOYrjweKB-zNNAijrQUqpeiTgOessyYSDApTMpqoWnG0EehFaepAY2R4iyULNSCtqGRFzmeA5GxoBnv2btCk2dlsZYYcamV0iaT0OoCWtZg00XVKWO6sdXlH-tXcGiPxZHjwmtomFfHG9hTn6tZubx15_sFxj2pkw
linkProvider IEEE
linkToHtml http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV09T8MwED2h8r0ApYhvPCC2tEkcx86IEFUr2qoqAVViqBzbkTo0qZqW34_tpBQGBrbIiqLobN3dO793B3CvIUFEOGGOjLhwAsGJw5TnOYTp2CGxYmEi7LAJOhiw8TgaVmJ1q4VRSlnymWqaR3uXL3OxMqWyls4N9BnCGqFvkyDw3VKutXa8BPuY7sFD1UazNeIaiRvdnMmaXUY1GPRJc_2NX9NUbDBpH_3zN46hsZHloeF3wDmBLZXV4fBHR8E67FpGpyhO4eMtK1Zz4wkKJVGcz_LXxxHqzrT_QAZzbjrHoric1YPepwujJkH9imSI-psCIirJemV9rwFx-zl-6jjVJAVnGuGlQ33hqTTxJYtCyZlwEyUTnyYk1SZihDOdtBpwmhLlKSYFxYmGjaGgJOAkkAyfQS3LM3UOiEcMp9Q1bwU600ojyVVIuRRC6lxCigtoGINN5mWvjMnaVpd_rN_Bfifu9ya97uDlCg7MFlmqXHANNW0GdQM74nM5LRa3dq-_AMIsrNo
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook&rft.genre=proceeding&rft.title=Proceedings+of+the+IEEE+National+Radar+Conference+%281996%29&rft.atitle=Unsupervised+TomoSAR+Image+Reconstruction+Through+Virtual+Multiple+Measurement+Explorations&rft.au=Liu%2C+Liang&rft.au=Zeng%2C+Tianjiao&rft.au=Wang%2C+Mou&rft.au=Shi%2C+Jun&rft.date=2025-10-04&rft.pub=IEEE&rft.eissn=2375-5318&rft.spage=321&rft.epage=326&rft_id=info:doi/10.1109%2FRadarConf2559087.2025.11205130&rft.externalDocID=11205130