SLIC Superpixel-Based <i>l</i><sub>2,1</sub>-Norm Robust Principal Component Analysis for Hyperspectral Image Classification
Hyperspectral Images (HSIs) contain enriched information due to the presence of various bands, which have gained attention for the past few decades. However, explosive growth in HSIs’ scale and dimensions causes “Curse of dimensionality„ and “Hughes phenomenon...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2019-01-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/19/3/479 |
| _version_ | 1811304939156144128 |
|---|---|
| author | Baokai Zu Kewen Xia Tiejun Li Ziping He Yafang Li Jingzhong Hou Wei Du |
| author_facet | Baokai Zu Kewen Xia Tiejun Li Ziping He Yafang Li Jingzhong Hou Wei Du |
| author_sort | Baokai Zu |
| collection | DOAJ |
| description | Hyperspectral Images (HSIs) contain enriched information due to the presence of various bands, which have gained attention for the past few decades. However, explosive growth in HSIs’ scale and dimensions causes “Curse of dimensionality„ and “Hughes phenomenon„. Dimensionality reduction has become an important means to overcome the “Curse of dimensionality„. In hyperspectral images, labeled samples are more difficult to collect because they require many labor and material resources. Semi-supervised dimensionality reduction is very important in mining high-dimensional data due to the lack of costly-labeled samples. The promotion of the supervised dimensionality reduction method to the semi-supervised method is mostly done by graph, which is a powerful tool for characterizing data relationships and manifold exploration. To take advantage of the spatial information of data, we put forward a novel graph construction method for semi-supervised learning, called SLIC Superpixel-based <inline-formula> <math display="inline"> <semantics> <msub> <mi>l</mi> <mrow> <mn>2</mn> <mo>,</mo> <mn>1</mn> </mrow> </msub> </semantics> </math> </inline-formula>-norm Robust Principal Component Analysis (SURPCA<sub>2,1</sub>), which integrates superpixel segmentation method Simple Linear Iterative Clustering (SLIC) into Low-rank Decomposition. First, the SLIC algorithm is adopted to obtain the spatial homogeneous regions of HSI. Then, the <inline-formula> <math display="inline"> <semantics> <msub> <mi>l</mi> <mrow> <mn>2</mn> <mo>,</mo> <mn>1</mn> </mrow> </msub> </semantics> </math> </inline-formula>-norm RPCA is exploited in each superpixel area, which captures the global information of homogeneous regions and preserves spectral subspace segmentation of HSIs very well. Therefore, we have explored the spatial and spectral information of hyperspectral image simultaneously by combining superpixel segmentation with RPCA. Finally, a semi-supervised dimensionality reduction framework based on SURPCA<sub>2,1</sub> graph is used for feature extraction task. Extensive experiments on multiple HSIs showed that the proposed spectral-spatial SURPCA<sub>2,1</sub> is always comparable to other compared graphs with few labeled samples. |
| first_indexed | 2024-04-13T08:16:24Z |
| format | Article |
| id | doaj.art-f6ee81cb74bb4e28a2578ad3e76b94b9 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-04-13T08:16:24Z |
| publishDate | 2019-01-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-f6ee81cb74bb4e28a2578ad3e76b94b92022-12-22T02:54:46ZengMDPI AGSensors1424-82202019-01-0119347910.3390/s19030479s19030479SLIC Superpixel-Based <i>l</i><sub>2,1</sub>-Norm Robust Principal Component Analysis for Hyperspectral Image ClassificationBaokai Zu0Kewen Xia1Tiejun Li2Ziping He3Yafang Li4Jingzhong Hou5Wei Du6School of Electronics and Information Engineering, Hebei University of Technology, Tianjin 300401, ChinaSchool of Electronics and Information Engineering, Hebei University of Technology, Tianjin 300401, ChinaSchool of Mechanical Engineering, Hebei University of Technology, Tianjin 300401, ChinaSchool of Electronics and Information Engineering, Hebei University of Technology, Tianjin 300401, ChinaFaculty of Information Technology, Beijing University of Technology, Beijing 100124, ChinaSchool of Electronics and Information Engineering, Hebei University of Technology, Tianjin 300401, ChinaCollege of Resources and Environment, Huazhong Agricultural University, Wuhan 430070, ChinaHyperspectral Images (HSIs) contain enriched information due to the presence of various bands, which have gained attention for the past few decades. However, explosive growth in HSIs’ scale and dimensions causes “Curse of dimensionality„ and “Hughes phenomenon„. Dimensionality reduction has become an important means to overcome the “Curse of dimensionality„. In hyperspectral images, labeled samples are more difficult to collect because they require many labor and material resources. Semi-supervised dimensionality reduction is very important in mining high-dimensional data due to the lack of costly-labeled samples. The promotion of the supervised dimensionality reduction method to the semi-supervised method is mostly done by graph, which is a powerful tool for characterizing data relationships and manifold exploration. To take advantage of the spatial information of data, we put forward a novel graph construction method for semi-supervised learning, called SLIC Superpixel-based <inline-formula> <math display="inline"> <semantics> <msub> <mi>l</mi> <mrow> <mn>2</mn> <mo>,</mo> <mn>1</mn> </mrow> </msub> </semantics> </math> </inline-formula>-norm Robust Principal Component Analysis (SURPCA<sub>2,1</sub>), which integrates superpixel segmentation method Simple Linear Iterative Clustering (SLIC) into Low-rank Decomposition. First, the SLIC algorithm is adopted to obtain the spatial homogeneous regions of HSI. Then, the <inline-formula> <math display="inline"> <semantics> <msub> <mi>l</mi> <mrow> <mn>2</mn> <mo>,</mo> <mn>1</mn> </mrow> </msub> </semantics> </math> </inline-formula>-norm RPCA is exploited in each superpixel area, which captures the global information of homogeneous regions and preserves spectral subspace segmentation of HSIs very well. Therefore, we have explored the spatial and spectral information of hyperspectral image simultaneously by combining superpixel segmentation with RPCA. Finally, a semi-supervised dimensionality reduction framework based on SURPCA<sub>2,1</sub> graph is used for feature extraction task. Extensive experiments on multiple HSIs showed that the proposed spectral-spatial SURPCA<sub>2,1</sub> is always comparable to other compared graphs with few labeled samples.https://www.mdpi.com/1424-8220/19/3/479Hyperspectral ImageRobust Principal Component Analysis (RPCA)Simple Linear Iterative Clustering (SLIC)superpixel segmentation |
| spellingShingle | Baokai Zu Kewen Xia Tiejun Li Ziping He Yafang Li Jingzhong Hou Wei Du SLIC Superpixel-Based <i>l</i><sub>2,1</sub>-Norm Robust Principal Component Analysis for Hyperspectral Image Classification Sensors Hyperspectral Image Robust Principal Component Analysis (RPCA) Simple Linear Iterative Clustering (SLIC) superpixel segmentation |
| title | SLIC Superpixel-Based <i>l</i><sub>2,1</sub>-Norm Robust Principal Component Analysis for Hyperspectral Image Classification |
| title_full | SLIC Superpixel-Based <i>l</i><sub>2,1</sub>-Norm Robust Principal Component Analysis for Hyperspectral Image Classification |
| title_fullStr | SLIC Superpixel-Based <i>l</i><sub>2,1</sub>-Norm Robust Principal Component Analysis for Hyperspectral Image Classification |
| title_full_unstemmed | SLIC Superpixel-Based <i>l</i><sub>2,1</sub>-Norm Robust Principal Component Analysis for Hyperspectral Image Classification |
| title_short | SLIC Superpixel-Based <i>l</i><sub>2,1</sub>-Norm Robust Principal Component Analysis for Hyperspectral Image Classification |
| title_sort | slic superpixel based i l i sub 2 1 sub norm robust principal component analysis for hyperspectral image classification |
| topic | Hyperspectral Image Robust Principal Component Analysis (RPCA) Simple Linear Iterative Clustering (SLIC) superpixel segmentation |
| url | https://www.mdpi.com/1424-8220/19/3/479 |
| work_keys_str_mv | AT baokaizu slicsuperpixelbasedilisub21subnormrobustprincipalcomponentanalysisforhyperspectralimageclassification AT kewenxia slicsuperpixelbasedilisub21subnormrobustprincipalcomponentanalysisforhyperspectralimageclassification AT tiejunli slicsuperpixelbasedilisub21subnormrobustprincipalcomponentanalysisforhyperspectralimageclassification AT zipinghe slicsuperpixelbasedilisub21subnormrobustprincipalcomponentanalysisforhyperspectralimageclassification AT yafangli slicsuperpixelbasedilisub21subnormrobustprincipalcomponentanalysisforhyperspectralimageclassification AT jingzhonghou slicsuperpixelbasedilisub21subnormrobustprincipalcomponentanalysisforhyperspectralimageclassification AT weidu slicsuperpixelbasedilisub21subnormrobustprincipalcomponentanalysisforhyperspectralimageclassification |