Journal article 132 views
Farmland Segmentation in Landsat 8 Satellite Images Using Deep Learning and Conditional Generative Adversarial Networks
Remote Sensing, Volume: 16, Issue: 5, Start page: 823
Swansea University Author:
Sara Sharifzadeh
Full text not available from this repository: check for access using links below.
DOI (Published version): 10.3390/rs16050823
Abstract
Leveraging mid-resolution satellite images such as Landsat 8 for accurate farmland segmentation and land change monitoring is crucial for agricultural management, yet is hindered by the scarcity of labelled data for the training of supervised deep learning pipelines. The particular focus of this stu...
Published in: | Remote Sensing |
---|---|
ISSN: | 2072-4292 |
Published: |
MDPI AG
2024
|
Online Access: |
Check full text
|
URI: | https://cronfa.swan.ac.uk/Record/cronfa65720 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
first_indexed |
2024-04-25T21:22:35Z |
---|---|
last_indexed |
2024-04-25T21:22:35Z |
id |
cronfa65720 |
recordtype |
SURis |
fullrecord |
<?xml version="1.0" encoding="utf-8"?><rfc1807 xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xsd="http://www.w3.org/2001/XMLSchema"><bib-version>v2</bib-version><id>65720</id><entry>2024-03-01</entry><title>Farmland Segmentation in Landsat 8 Satellite Images Using Deep Learning and Conditional Generative Adversarial Networks</title><swanseaauthors><author><sid>a4e15f304398ecee3f28c7faec69c1b0</sid><ORCID>0000-0003-4621-2917</ORCID><firstname>Sara</firstname><surname>Sharifzadeh</surname><name>Sara Sharifzadeh</name><active>true</active><ethesisStudent>false</ethesisStudent></author></swanseaauthors><date>2024-03-01</date><deptcode>SCS</deptcode><abstract>Leveraging mid-resolution satellite images such as Landsat 8 for accurate farmland segmentation and land change monitoring is crucial for agricultural management, yet is hindered by the scarcity of labelled data for the training of supervised deep learning pipelines. The particular focus of this study is on addressing the scarcity of labelled images. This paper introduces several contributions, including a systematic satellite image data augmentation approach that aims to maintain data population consistency during model training, thus mitigating performance degradation. To alleviate the labour-intensive task of pixel-wise image labelling, we present a novel application of a modified conditional generative adversarial network (CGAN) to generate artificial satellite images and corresponding farm labels. Additionally, we scrutinize the role of spectral bands in satellite image segmentation and compare two prominent semantic segmentation models, U-Net and DeepLabV3+, with diverse backbone structures. Our empirical findings demonstrate that augmenting the dataset with up to 22.85% artificial samples significantly enhances the model performance. Notably, the U-Net model, employing standard convolution, outperforms the DeepLabV3+ models with atrous convolution, achieving a segmentation accuracy of 86.92% on the test data.</abstract><type>Journal Article</type><journal>Remote Sensing</journal><volume>16</volume><journalNumber>5</journalNumber><paginationStart>823</paginationStart><paginationEnd/><publisher>MDPI AG</publisher><placeOfPublication/><isbnPrint/><isbnElectronic/><issnPrint/><issnElectronic>2072-4292</issnElectronic><keywords>farm segmentation; deep learning; semantic segmentation; NDVI; CGANs</keywords><publishedDay>27</publishedDay><publishedMonth>2</publishedMonth><publishedYear>2024</publishedYear><publishedDate>2024-02-27</publishedDate><doi>10.3390/rs16050823</doi><url/><notes/><college>COLLEGE NANME</college><department>Computer Science</department><CollegeCode>COLLEGE CODE</CollegeCode><DepartmentCode>SCS</DepartmentCode><institution>Swansea University</institution><apcterm>Another institution paid the OA fee</apcterm><funders>This research was generously supported by the Global Challenges Research Fund (GCRF) call 2020–2021.</funders><projectreference/><lastEdited>2024-04-25T22:24:26.5809043</lastEdited><Created>2024-03-01T11:21:33.7285952</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>Shruti</firstname><surname>Nair</surname><orcid>0000-0002-1763-7062</orcid><order>1</order></author><author><firstname>Sara</firstname><surname>Sharifzadeh</surname><orcid>0000-0003-4621-2917</orcid><order>2</order></author><author><firstname>Vasile</firstname><surname>Palade</surname><orcid>0000-0002-6768-8394</orcid><order>3</order></author></authors><documents/><OutputDurs/></rfc1807> |
spelling |
v2 65720 2024-03-01 Farmland Segmentation in Landsat 8 Satellite Images Using Deep Learning and Conditional Generative Adversarial Networks a4e15f304398ecee3f28c7faec69c1b0 0000-0003-4621-2917 Sara Sharifzadeh Sara Sharifzadeh true false 2024-03-01 SCS Leveraging mid-resolution satellite images such as Landsat 8 for accurate farmland segmentation and land change monitoring is crucial for agricultural management, yet is hindered by the scarcity of labelled data for the training of supervised deep learning pipelines. The particular focus of this study is on addressing the scarcity of labelled images. This paper introduces several contributions, including a systematic satellite image data augmentation approach that aims to maintain data population consistency during model training, thus mitigating performance degradation. To alleviate the labour-intensive task of pixel-wise image labelling, we present a novel application of a modified conditional generative adversarial network (CGAN) to generate artificial satellite images and corresponding farm labels. Additionally, we scrutinize the role of spectral bands in satellite image segmentation and compare two prominent semantic segmentation models, U-Net and DeepLabV3+, with diverse backbone structures. Our empirical findings demonstrate that augmenting the dataset with up to 22.85% artificial samples significantly enhances the model performance. Notably, the U-Net model, employing standard convolution, outperforms the DeepLabV3+ models with atrous convolution, achieving a segmentation accuracy of 86.92% on the test data. Journal Article Remote Sensing 16 5 823 MDPI AG 2072-4292 farm segmentation; deep learning; semantic segmentation; NDVI; CGANs 27 2 2024 2024-02-27 10.3390/rs16050823 COLLEGE NANME Computer Science COLLEGE CODE SCS Swansea University Another institution paid the OA fee This research was generously supported by the Global Challenges Research Fund (GCRF) call 2020–2021. 2024-04-25T22:24:26.5809043 2024-03-01T11:21:33.7285952 Faculty of Science and Engineering School of Mathematics and Computer Science - Computer Science Shruti Nair 0000-0002-1763-7062 1 Sara Sharifzadeh 0000-0003-4621-2917 2 Vasile Palade 0000-0002-6768-8394 3 |
title |
Farmland Segmentation in Landsat 8 Satellite Images Using Deep Learning and Conditional Generative Adversarial Networks |
spellingShingle |
Farmland Segmentation in Landsat 8 Satellite Images Using Deep Learning and Conditional Generative Adversarial Networks Sara Sharifzadeh |
title_short |
Farmland Segmentation in Landsat 8 Satellite Images Using Deep Learning and Conditional Generative Adversarial Networks |
title_full |
Farmland Segmentation in Landsat 8 Satellite Images Using Deep Learning and Conditional Generative Adversarial Networks |
title_fullStr |
Farmland Segmentation in Landsat 8 Satellite Images Using Deep Learning and Conditional Generative Adversarial Networks |
title_full_unstemmed |
Farmland Segmentation in Landsat 8 Satellite Images Using Deep Learning and Conditional Generative Adversarial Networks |
title_sort |
Farmland Segmentation in Landsat 8 Satellite Images Using Deep Learning and Conditional Generative Adversarial Networks |
author_id_str_mv |
a4e15f304398ecee3f28c7faec69c1b0 |
author_id_fullname_str_mv |
a4e15f304398ecee3f28c7faec69c1b0_***_Sara Sharifzadeh |
author |
Sara Sharifzadeh |
author2 |
Shruti Nair Sara Sharifzadeh Vasile Palade |
format |
Journal article |
container_title |
Remote Sensing |
container_volume |
16 |
container_issue |
5 |
container_start_page |
823 |
publishDate |
2024 |
institution |
Swansea University |
issn |
2072-4292 |
doi_str_mv |
10.3390/rs16050823 |
publisher |
MDPI AG |
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 |
0 |
active_str |
0 |
description |
Leveraging mid-resolution satellite images such as Landsat 8 for accurate farmland segmentation and land change monitoring is crucial for agricultural management, yet is hindered by the scarcity of labelled data for the training of supervised deep learning pipelines. The particular focus of this study is on addressing the scarcity of labelled images. This paper introduces several contributions, including a systematic satellite image data augmentation approach that aims to maintain data population consistency during model training, thus mitigating performance degradation. To alleviate the labour-intensive task of pixel-wise image labelling, we present a novel application of a modified conditional generative adversarial network (CGAN) to generate artificial satellite images and corresponding farm labels. Additionally, we scrutinize the role of spectral bands in satellite image segmentation and compare two prominent semantic segmentation models, U-Net and DeepLabV3+, with diverse backbone structures. Our empirical findings demonstrate that augmenting the dataset with up to 22.85% artificial samples significantly enhances the model performance. Notably, the U-Net model, employing standard convolution, outperforms the DeepLabV3+ models with atrous convolution, achieving a segmentation accuracy of 86.92% on the test data. |
published_date |
2024-02-27T22:24:27Z |
_version_ |
1797343430245875712 |
score |
11.021648 |