Local and Global Spatial Information for Land Cover Semisupervised Classification of Complex Polarimetric SAR Data
Each of the three satellites constituting the RADARSAT Constellation Mission (RCM) provides compact polarimetric synthetic aperture radar (CP SAR) data. The complex CP data have similar properties to the complex quad polarimetric (QP) data provided by prior RADARSAT missions. In this article, a land...
Main Authors: | , , |
---|---|
Format: | Article |
Language: | English |
Published: |
IEEE
2023-01-01
|
Series: | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
Subjects: | |
Online Access: | https://ieeexplore.ieee.org/document/10093952/ |
_version_ | 1797835817441296384 |
---|---|
author | Mohsen Ghanbari Linlin Xu David A. Clausi |
author_facet | Mohsen Ghanbari Linlin Xu David A. Clausi |
author_sort | Mohsen Ghanbari |
collection | DOAJ |
description | Each of the three satellites constituting the RADARSAT Constellation Mission (RCM) provides compact polarimetric synthetic aperture radar (CP SAR) data. The complex CP data have similar properties to the complex quad polarimetric (QP) data provided by prior RADARSAT missions. In this article, a land cover classification method using spatial information is designed based on the statistical characteristics of the complex CP and QP SAR data. First, the local spatial dependency among pixels is captured by superpixels. Second, a graph is constructed on the superpixels to model the global spatial dependency among superpixels. The land cover classification image with land cover type labels is then estimated by propagating labels from the few labeled superpixels to the unlabeled superpixels. Classification of two RCM complex CP and QP scenes demonstrates that the proposed method, with few labeled pixels, provides much higher classification accuracy than methods that do not exploit global spatial dependency. |
first_indexed | 2024-04-09T14:58:35Z |
format | Article |
id | doaj.art-65e15e06adfe4d3c8d933e80b06037dc |
institution | Directory Open Access Journal |
issn | 2151-1535 |
language | English |
last_indexed | 2024-04-09T14:58:35Z |
publishDate | 2023-01-01 |
publisher | IEEE |
record_format | Article |
series | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
spelling | doaj.art-65e15e06adfe4d3c8d933e80b06037dc2023-05-01T23:00:17ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing2151-15352023-01-01163892390410.1109/JSTARS.2023.326445210093952Local and Global Spatial Information for Land Cover Semisupervised Classification of Complex Polarimetric SAR DataMohsen Ghanbari0https://orcid.org/0000-0003-2793-982XLinlin Xu1https://orcid.org/0000-0002-3488-5199David A. Clausi2https://orcid.org/0000-0002-6383-0875Vision and Image Processing (VIP) Research Group, Department of Systems Design Engineering, University of Waterloo, Waterloo, ON, CanadaVision and Image Processing (VIP) Research Group, Department of Systems Design Engineering, University of Waterloo, Waterloo, ON, CanadaVision and Image Processing (VIP) Research Group, Department of Systems Design Engineering, University of Waterloo, Waterloo, ON, CanadaEach of the three satellites constituting the RADARSAT Constellation Mission (RCM) provides compact polarimetric synthetic aperture radar (CP SAR) data. The complex CP data have similar properties to the complex quad polarimetric (QP) data provided by prior RADARSAT missions. In this article, a land cover classification method using spatial information is designed based on the statistical characteristics of the complex CP and QP SAR data. First, the local spatial dependency among pixels is captured by superpixels. Second, a graph is constructed on the superpixels to model the global spatial dependency among superpixels. The land cover classification image with land cover type labels is then estimated by propagating labels from the few labeled superpixels to the unlabeled superpixels. Classification of two RCM complex CP and QP scenes demonstrates that the proposed method, with few labeled pixels, provides much higher classification accuracy than methods that do not exploit global spatial dependency.https://ieeexplore.ieee.org/document/10093952/Compact polarimetrygraph-based learningland cover classificationmultilook complexsemisupervisedsuperpixel |
spellingShingle | Mohsen Ghanbari Linlin Xu David A. Clausi Local and Global Spatial Information for Land Cover Semisupervised Classification of Complex Polarimetric SAR Data IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing Compact polarimetry graph-based learning land cover classification multilook complex semisupervised superpixel |
title | Local and Global Spatial Information for Land Cover Semisupervised Classification of Complex Polarimetric SAR Data |
title_full | Local and Global Spatial Information for Land Cover Semisupervised Classification of Complex Polarimetric SAR Data |
title_fullStr | Local and Global Spatial Information for Land Cover Semisupervised Classification of Complex Polarimetric SAR Data |
title_full_unstemmed | Local and Global Spatial Information for Land Cover Semisupervised Classification of Complex Polarimetric SAR Data |
title_short | Local and Global Spatial Information for Land Cover Semisupervised Classification of Complex Polarimetric SAR Data |
title_sort | local and global spatial information for land cover semisupervised classification of complex polarimetric sar data |
topic | Compact polarimetry graph-based learning land cover classification multilook complex semisupervised superpixel |
url | https://ieeexplore.ieee.org/document/10093952/ |
work_keys_str_mv | AT mohsenghanbari localandglobalspatialinformationforlandcoversemisupervisedclassificationofcomplexpolarimetricsardata AT linlinxu localandglobalspatialinformationforlandcoversemisupervisedclassificationofcomplexpolarimetricsardata AT davidaclausi localandglobalspatialinformationforlandcoversemisupervisedclassificationofcomplexpolarimetricsardata |