Semisupervised Hyperspectral Image Classification via Superpixel-Based Graph Regularization With Local and Nonlocal Features
Although a graph-based semisupervised learning (SSL) approach can utilize limited numbers of labeled samples for hyperspectral image (HSI) classification, it is difficult to use the large amount of pixels in an HSI to construct a large-scale graph. In this article, we therefore propose a superpixel-...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2022-01-01
|
| Series: | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/9832660/ |
| _version_ | 1828433130540761088 |
|---|---|
| author | Longshan Yang Junhuan Peng Yuebin Wang Linlin Xu Weiwei Zhu |
| author_facet | Longshan Yang Junhuan Peng Yuebin Wang Linlin Xu Weiwei Zhu |
| author_sort | Longshan Yang |
| collection | DOAJ |
| description | Although a graph-based semisupervised learning (SSL) approach can utilize limited numbers of labeled samples for hyperspectral image (HSI) classification, it is difficult to use the large amount of pixels in an HSI to construct a large-scale graph. In this article, we therefore propose a superpixel-based graph model for HSI classification, using anchor graph regularization to improve the extraction of local and nonlocal spatial information. To avoid the large-scale graph problem, the superpixel-based graph model constructs a scalable anchor graph using a small number of anchor points for graph-based SSL. The proposed approach consists of three key components: First, local and nonlocal features are extracted from the principal components of an HSI using the locally grouped order pattern and the nonlocal binary pattern approaches. Second, the extracted features are combined with the original spectral information and used as input to an improved simple line iteration clustering method to obtain superpixels, the centers of which are used as anchors in the anchor graph approach. Third, we use superpixel-based graph regularization to model the global information among the superpixels for better predictions of the label of each unlabeled pixel sample. Experimental results demonstrate that the proposed approach outperforms other state-of-the-art semisupervised HSI classification approaches for cases with a limited quantity of labeled samples. |
| first_indexed | 2024-12-10T18:30:38Z |
| format | Article |
| id | doaj.art-8b356ac688704b5faa6cb1e9d601bd2e |
| institution | Directory Open Access Journal |
| issn | 2151-1535 |
| language | English |
| last_indexed | 2024-12-10T18:30:38Z |
| publishDate | 2022-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| spelling | doaj.art-8b356ac688704b5faa6cb1e9d601bd2e2022-12-22T01:37:57ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing2151-15352022-01-01156645665810.1109/JSTARS.2022.31916929832660Semisupervised Hyperspectral Image Classification via Superpixel-Based Graph Regularization With Local and Nonlocal FeaturesLongshan Yang0https://orcid.org/0000-0002-3509-2123Junhuan Peng1Yuebin Wang2https://orcid.org/0000-0002-6978-4558Linlin Xu3https://orcid.org/0000-0002-3488-5199Weiwei Zhu4School of Land Science and Technology, China University of Geosciences, Beijing, ChinaSchool of Land Science and Technology, China University of Geosciences, Beijing, ChinaSchool of Land Science and Technology, China University of Geosciences, Beijing, ChinaSchool of Land Science and Technology, China University of Geosciences, Beijing, ChinaSatellite Application Center, Ministry of Ecology and Environment, Beijing, ChinaAlthough a graph-based semisupervised learning (SSL) approach can utilize limited numbers of labeled samples for hyperspectral image (HSI) classification, it is difficult to use the large amount of pixels in an HSI to construct a large-scale graph. In this article, we therefore propose a superpixel-based graph model for HSI classification, using anchor graph regularization to improve the extraction of local and nonlocal spatial information. To avoid the large-scale graph problem, the superpixel-based graph model constructs a scalable anchor graph using a small number of anchor points for graph-based SSL. The proposed approach consists of three key components: First, local and nonlocal features are extracted from the principal components of an HSI using the locally grouped order pattern and the nonlocal binary pattern approaches. Second, the extracted features are combined with the original spectral information and used as input to an improved simple line iteration clustering method to obtain superpixels, the centers of which are used as anchors in the anchor graph approach. Third, we use superpixel-based graph regularization to model the global information among the superpixels for better predictions of the label of each unlabeled pixel sample. Experimental results demonstrate that the proposed approach outperforms other state-of-the-art semisupervised HSI classification approaches for cases with a limited quantity of labeled samples.https://ieeexplore.ieee.org/document/9832660/Anchor graphfeature-extractionhyperspectral imagesemisupervised classificationsuperpixel |
| spellingShingle | Longshan Yang Junhuan Peng Yuebin Wang Linlin Xu Weiwei Zhu Semisupervised Hyperspectral Image Classification via Superpixel-Based Graph Regularization With Local and Nonlocal Features IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing Anchor graph feature-extraction hyperspectral image semisupervised classification superpixel |
| title | Semisupervised Hyperspectral Image Classification via Superpixel-Based Graph Regularization With Local and Nonlocal Features |
| title_full | Semisupervised Hyperspectral Image Classification via Superpixel-Based Graph Regularization With Local and Nonlocal Features |
| title_fullStr | Semisupervised Hyperspectral Image Classification via Superpixel-Based Graph Regularization With Local and Nonlocal Features |
| title_full_unstemmed | Semisupervised Hyperspectral Image Classification via Superpixel-Based Graph Regularization With Local and Nonlocal Features |
| title_short | Semisupervised Hyperspectral Image Classification via Superpixel-Based Graph Regularization With Local and Nonlocal Features |
| title_sort | semisupervised hyperspectral image classification via superpixel based graph regularization with local and nonlocal features |
| topic | Anchor graph feature-extraction hyperspectral image semisupervised classification superpixel |
| url | https://ieeexplore.ieee.org/document/9832660/ |
| work_keys_str_mv | AT longshanyang semisupervisedhyperspectralimageclassificationviasuperpixelbasedgraphregularizationwithlocalandnonlocalfeatures AT junhuanpeng semisupervisedhyperspectralimageclassificationviasuperpixelbasedgraphregularizationwithlocalandnonlocalfeatures AT yuebinwang semisupervisedhyperspectralimageclassificationviasuperpixelbasedgraphregularizationwithlocalandnonlocalfeatures AT linlinxu semisupervisedhyperspectralimageclassificationviasuperpixelbasedgraphregularizationwithlocalandnonlocalfeatures AT weiweizhu semisupervisedhyperspectralimageclassificationviasuperpixelbasedgraphregularizationwithlocalandnonlocalfeatures |