No Cover Image

E-Thesis 17 views 6 downloads

Mass labelling Remote Sensing Images: A visualisation and machine learning combinatory approach / TULSI PATEL

Swansea University Author: TULSI PATEL

  • 2025_Patel_T.final.71617.pdf

    PDF | E-Thesis – open access

    Copyright: the author, Tulsi Patel, 2026. Distributed under the terms of a Creative Commons Attribution 4.0 License (CC BY 4.0).

    Download (34.11MB)

DOI (Published version): 10.23889/SUThesis.71617

Abstract

Deep learning and computer vision have made significant advancements in the field of the automated analysis of Remote Sensing imagery. However, in order to maximise the utility of satellite derived Remote sensing imagery, numerous challenges must be addressed. Firstly, the complexity of the Earth’s su...

Full description

Published: Swansea 2025
Institution: Swansea University
Degree level: Doctoral
Degree name: Ph.D
Supervisor: Jones, M. W.
URI: https://cronfa.swan.ac.uk/Record/cronfa71617
first_indexed 2026-03-12T12:42:30Z
last_indexed 2026-03-13T05:25:17Z
id cronfa71617
recordtype RisThesis
fullrecord <?xml version="1.0"?><rfc1807><datestamp>2026-03-12T12:49:40.5370363</datestamp><bib-version>v2</bib-version><id>71617</id><entry>2026-03-12</entry><title>Mass labelling Remote Sensing Images: A visualisation and machine learning combinatory approach</title><swanseaauthors><author><sid>5ce9653007f84ed824ef8e6b28b7fc2c</sid><firstname>TULSI</firstname><surname>PATEL</surname><name>TULSI PATEL</name><active>true</active><ethesisStudent>false</ethesisStudent></author></swanseaauthors><date>2026-03-12</date><abstract>Deep learning and computer vision have made signi&#xFB01;cant advancements in the &#xFB01;eld of the automated analysis of Remote Sensing imagery. However, in order to maximise the utility of satellite derived Remote sensing imagery, numerous challenges must be addressed. Firstly, the complexity of the Earth&#x2019;s surface, in conjunction with its atmosphere, introduces variation within data, requiring expert knowledge to interpret, label and ensure accuracy. Secondly, a large volume of imagery data is generated every day from a single sensing instrument and there are multiple instruments continuously monitoring the earth&#x2019;s surface. Only a very small percentage of data has been accurately labelled, and that which has been is targeted towards a speci&#xFB01;c domains and of limited temporal resolution. In this thesis, we identify the need for and explore a new remote image labelling framework. We utilise visualisation, unsupervised ma-chine learning and remote sensing methodologies to enable users to create large-scale, labelled remote sensing datasets. Our approach combines three main components - Deep convolutional auto-encoders, manifold learning and an interactive data labelling application. The deep convolutional auto-encoder learns a condensed and informative representations satellite imagery.The manifold learning condenses this representation down into two dimensions, which then supports visualisation techniques that help to convey patterns and representations to a labeller.Graph neural networks are then introduced to further enhance the spatial encoding of geographical features within imagery, and the use of super-pixel representations allow users to create full segmentation labels. Throughout the thesis, we present an evolving visualisation application in which we explore the feature encodings provided by our deep learning pipelines. We introduce a novel methodology for branching and merging datasets, providing a more &#xFB01;ne-grained and expressive labelling experience for users. Overall, this thesis presents novel and innovative methods for labelling remote sensing images, using techniques from visualisation, computer vision and the remote sensing domains.</abstract><type>E-Thesis</type><journal/><volume/><journalNumber/><paginationStart/><paginationEnd/><publisher/><placeOfPublication>Swansea</placeOfPublication><isbnPrint/><isbnElectronic/><issnPrint/><issnElectronic/><keywords>Remote Sensing, Manifold Learning, Visualisation, Labelling</keywords><publishedDay>15</publishedDay><publishedMonth>12</publishedMonth><publishedYear>2025</publishedYear><publishedDate>2025-12-15</publishedDate><doi>10.23889/SUThesis.71617</doi><url/><notes/><college>COLLEGE NANME</college><CollegeCode>COLLEGE CODE</CollegeCode><institution>Swansea University</institution><supervisor>Jones, M. W.</supervisor><degreelevel>Doctoral</degreelevel><degreename>Ph.D</degreename><degreesponsorsfunders>EPSRC Doctoral Training Grant</degreesponsorsfunders><apcterm/><funders>EPSRC Doctoral Training Grant</funders><projectreference/><lastEdited>2026-03-12T12:49:40.5370363</lastEdited><Created>2026-03-12T12:36:51.6661860</Created><path><level id="1">Faculty of Science and Engineering</level><level id="2">School of Mathematics and Computer Science - Computer Science</level></path><authors><author><firstname>TULSI</firstname><surname>PATEL</surname><order>1</order></author></authors><documents><document><filename>71617__36394__e10302e5e137495e8237e2eca34caf29.pdf</filename><originalFilename>2025_Patel_T.final.71617.pdf</originalFilename><uploaded>2026-03-12T12:44:20.7977558</uploaded><type>Output</type><contentLength>35764987</contentLength><contentType>application/pdf</contentType><version>E-Thesis &#x2013; open access</version><cronfaStatus>true</cronfaStatus><documentNotes>Copyright: the author, Tulsi Patel, 2026. Distributed under the terms of a Creative Commons Attribution 4.0 License (CC BY 4.0).</documentNotes><copyrightCorrect>true</copyrightCorrect><language>eng</language><licence>https://creativecommons.org/licenses/by/4.0/</licence></document></documents><OutputDurs/></rfc1807>
spelling 2026-03-12T12:49:40.5370363 v2 71617 2026-03-12 Mass labelling Remote Sensing Images: A visualisation and machine learning combinatory approach 5ce9653007f84ed824ef8e6b28b7fc2c TULSI PATEL TULSI PATEL true false 2026-03-12 Deep learning and computer vision have made significant advancements in the field of the automated analysis of Remote Sensing imagery. However, in order to maximise the utility of satellite derived Remote sensing imagery, numerous challenges must be addressed. Firstly, the complexity of the Earth’s surface, in conjunction with its atmosphere, introduces variation within data, requiring expert knowledge to interpret, label and ensure accuracy. Secondly, a large volume of imagery data is generated every day from a single sensing instrument and there are multiple instruments continuously monitoring the earth’s surface. Only a very small percentage of data has been accurately labelled, and that which has been is targeted towards a specific domains and of limited temporal resolution. In this thesis, we identify the need for and explore a new remote image labelling framework. We utilise visualisation, unsupervised ma-chine learning and remote sensing methodologies to enable users to create large-scale, labelled remote sensing datasets. Our approach combines three main components - Deep convolutional auto-encoders, manifold learning and an interactive data labelling application. The deep convolutional auto-encoder learns a condensed and informative representations satellite imagery.The manifold learning condenses this representation down into two dimensions, which then supports visualisation techniques that help to convey patterns and representations to a labeller.Graph neural networks are then introduced to further enhance the spatial encoding of geographical features within imagery, and the use of super-pixel representations allow users to create full segmentation labels. Throughout the thesis, we present an evolving visualisation application in which we explore the feature encodings provided by our deep learning pipelines. We introduce a novel methodology for branching and merging datasets, providing a more fine-grained and expressive labelling experience for users. Overall, this thesis presents novel and innovative methods for labelling remote sensing images, using techniques from visualisation, computer vision and the remote sensing domains. E-Thesis Swansea Remote Sensing, Manifold Learning, Visualisation, Labelling 15 12 2025 2025-12-15 10.23889/SUThesis.71617 COLLEGE NANME COLLEGE CODE Swansea University Jones, M. W. Doctoral Ph.D EPSRC Doctoral Training Grant EPSRC Doctoral Training Grant 2026-03-12T12:49:40.5370363 2026-03-12T12:36:51.6661860 Faculty of Science and Engineering School of Mathematics and Computer Science - Computer Science TULSI PATEL 1 71617__36394__e10302e5e137495e8237e2eca34caf29.pdf 2025_Patel_T.final.71617.pdf 2026-03-12T12:44:20.7977558 Output 35764987 application/pdf E-Thesis – open access true Copyright: the author, Tulsi Patel, 2026. Distributed under the terms of a Creative Commons Attribution 4.0 License (CC BY 4.0). true eng https://creativecommons.org/licenses/by/4.0/
title Mass labelling Remote Sensing Images: A visualisation and machine learning combinatory approach
spellingShingle Mass labelling Remote Sensing Images: A visualisation and machine learning combinatory approach
TULSI PATEL
title_short Mass labelling Remote Sensing Images: A visualisation and machine learning combinatory approach
title_full Mass labelling Remote Sensing Images: A visualisation and machine learning combinatory approach
title_fullStr Mass labelling Remote Sensing Images: A visualisation and machine learning combinatory approach
title_full_unstemmed Mass labelling Remote Sensing Images: A visualisation and machine learning combinatory approach
title_sort Mass labelling Remote Sensing Images: A visualisation and machine learning combinatory approach
author_id_str_mv 5ce9653007f84ed824ef8e6b28b7fc2c
author_id_fullname_str_mv 5ce9653007f84ed824ef8e6b28b7fc2c_***_TULSI PATEL
author TULSI PATEL
author2 TULSI PATEL
format E-Thesis
publishDate 2025
institution Swansea University
doi_str_mv 10.23889/SUThesis.71617
college_str Faculty of Science and Engineering
hierarchytype
hierarchy_top_id facultyofscienceandengineering
hierarchy_top_title Faculty of Science and Engineering
hierarchy_parent_id facultyofscienceandengineering
hierarchy_parent_title Faculty of Science and Engineering
department_str School of Mathematics and Computer Science - Computer Science{{{_:::_}}}Faculty of Science and Engineering{{{_:::_}}}School of Mathematics and Computer Science - Computer Science
document_store_str 1
active_str 0
description Deep learning and computer vision have made significant advancements in the field of the automated analysis of Remote Sensing imagery. However, in order to maximise the utility of satellite derived Remote sensing imagery, numerous challenges must be addressed. Firstly, the complexity of the Earth’s surface, in conjunction with its atmosphere, introduces variation within data, requiring expert knowledge to interpret, label and ensure accuracy. Secondly, a large volume of imagery data is generated every day from a single sensing instrument and there are multiple instruments continuously monitoring the earth’s surface. Only a very small percentage of data has been accurately labelled, and that which has been is targeted towards a specific domains and of limited temporal resolution. In this thesis, we identify the need for and explore a new remote image labelling framework. We utilise visualisation, unsupervised ma-chine learning and remote sensing methodologies to enable users to create large-scale, labelled remote sensing datasets. Our approach combines three main components - Deep convolutional auto-encoders, manifold learning and an interactive data labelling application. The deep convolutional auto-encoder learns a condensed and informative representations satellite imagery.The manifold learning condenses this representation down into two dimensions, which then supports visualisation techniques that help to convey patterns and representations to a labeller.Graph neural networks are then introduced to further enhance the spatial encoding of geographical features within imagery, and the use of super-pixel representations allow users to create full segmentation labels. Throughout the thesis, we present an evolving visualisation application in which we explore the feature encodings provided by our deep learning pipelines. We introduce a novel methodology for branching and merging datasets, providing a more fine-grained and expressive labelling experience for users. Overall, this thesis presents novel and innovative methods for labelling remote sensing images, using techniques from visualisation, computer vision and the remote sensing domains.
published_date 2025-12-15T05:25:17Z
_version_ 1859523200492240896
score 11.099629

Similar Items