Weighted Nonlocal Low-Rank Tensor Decomposition Method for Sparse Unmixing of Hyperspectral Images
The low spatial resolution of hyperspectral images leads to the coexistence of multiple ground objects in a single pixel (called mixed pixels). A large number of mixed pixels in a hyperspectral image hinders the subsequent analysis and application of the image. In order to solve this problem, a nove...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2020-01-01
|
| Series: | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/9035393/ |
| _version_ | 1830272353623343104 |
|---|---|
| author | Le Sun Feiyang Wu Tianming Zhan Wei Liu Jin Wang Byeungwoo Jeon |
| author_facet | Le Sun Feiyang Wu Tianming Zhan Wei Liu Jin Wang Byeungwoo Jeon |
| author_sort | Le Sun |
| collection | DOAJ |
| description | The low spatial resolution of hyperspectral images leads to the coexistence of multiple ground objects in a single pixel (called mixed pixels). A large number of mixed pixels in a hyperspectral image hinders the subsequent analysis and application of the image. In order to solve this problem, a novel sparse unmixing method, which considers highly similar patches in nonlocal regions of a hyperspectral image, is proposed in this article. This method exploits spectral correlation by using collaborative sparsity regularization and spatial information by employing total variation and weighted nonlocal low-rank tensor regularization. To effectively utilize the tensor decomposition, nonlocal similar patches are first grouped together. Then, these nonlocal patches are stacked to form a patch group tensor. Finally, weighted low-rank tensor regularization is enforced to constrain the patch group to obtain an estimated low-rank abundance image. Experiments on simulated and real hyperspectral datasets validated the superiority of the proposed method in better maintaining fine details and obtaining better unmixing results. |
| first_indexed | 2024-12-18T23:22:47Z |
| format | Article |
| id | doaj.art-23e454b949004073992ab8bf31943504 |
| institution | Directory Open Access Journal |
| issn | 2151-1535 |
| language | English |
| last_indexed | 2024-12-18T23:22:47Z |
| publishDate | 2020-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| spelling | doaj.art-23e454b949004073992ab8bf319435042022-12-21T20:47:53ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing2151-15352020-01-01131174118810.1109/JSTARS.2020.29805769035393Weighted Nonlocal Low-Rank Tensor Decomposition Method for Sparse Unmixing of Hyperspectral ImagesLe Sun0https://orcid.org/0000-0001-6465-8678Feiyang Wu1Tianming Zhan2https://orcid.org/0000-0001-5030-3032Wei Liu3https://orcid.org/0000-0001-8503-4063Jin Wang4https://orcid.org/0000-0001-5473-8738Byeungwoo Jeon5https://orcid.org/0000-0002-5650-2881School of Computer and Software, Nanjing University of Information Science and Technology (NUIST), Nanjing, ChinaSchool of Computer and Software, Nanjing University of Information Science and Technology, Nanjing, ChinaSchool of Information Engineering, Nanjing Audit University, Nanjing, ChinaSchool of Information and Engineering, Yangzhou University, Yangzhou, ChinaSchool of Computer and Communication Engineering, Changsha University of Science and Technology, Changsha, ChinaSchool of Electric and Electronic Engineering, Sungkyunkwan University, Seoul, South KoreaThe low spatial resolution of hyperspectral images leads to the coexistence of multiple ground objects in a single pixel (called mixed pixels). A large number of mixed pixels in a hyperspectral image hinders the subsequent analysis and application of the image. In order to solve this problem, a novel sparse unmixing method, which considers highly similar patches in nonlocal regions of a hyperspectral image, is proposed in this article. This method exploits spectral correlation by using collaborative sparsity regularization and spatial information by employing total variation and weighted nonlocal low-rank tensor regularization. To effectively utilize the tensor decomposition, nonlocal similar patches are first grouped together. Then, these nonlocal patches are stacked to form a patch group tensor. Finally, weighted low-rank tensor regularization is enforced to constrain the patch group to obtain an estimated low-rank abundance image. Experiments on simulated and real hyperspectral datasets validated the superiority of the proposed method in better maintaining fine details and obtaining better unmixing results.https://ieeexplore.ieee.org/document/9035393/Low-ranknonlocal similaritysparse unmixingtensor decomposition |
| spellingShingle | Le Sun Feiyang Wu Tianming Zhan Wei Liu Jin Wang Byeungwoo Jeon Weighted Nonlocal Low-Rank Tensor Decomposition Method for Sparse Unmixing of Hyperspectral Images IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing Low-rank nonlocal similarity sparse unmixing tensor decomposition |
| title | Weighted Nonlocal Low-Rank Tensor Decomposition Method for Sparse Unmixing of Hyperspectral Images |
| title_full | Weighted Nonlocal Low-Rank Tensor Decomposition Method for Sparse Unmixing of Hyperspectral Images |
| title_fullStr | Weighted Nonlocal Low-Rank Tensor Decomposition Method for Sparse Unmixing of Hyperspectral Images |
| title_full_unstemmed | Weighted Nonlocal Low-Rank Tensor Decomposition Method for Sparse Unmixing of Hyperspectral Images |
| title_short | Weighted Nonlocal Low-Rank Tensor Decomposition Method for Sparse Unmixing of Hyperspectral Images |
| title_sort | weighted nonlocal low rank tensor decomposition method for sparse unmixing of hyperspectral images |
| topic | Low-rank nonlocal similarity sparse unmixing tensor decomposition |
| url | https://ieeexplore.ieee.org/document/9035393/ |
| work_keys_str_mv | AT lesun weightednonlocallowranktensordecompositionmethodforsparseunmixingofhyperspectralimages AT feiyangwu weightednonlocallowranktensordecompositionmethodforsparseunmixingofhyperspectralimages AT tianmingzhan weightednonlocallowranktensordecompositionmethodforsparseunmixingofhyperspectralimages AT weiliu weightednonlocallowranktensordecompositionmethodforsparseunmixingofhyperspectralimages AT jinwang weightednonlocallowranktensordecompositionmethodforsparseunmixingofhyperspectralimages AT byeungwoojeon weightednonlocallowranktensordecompositionmethodforsparseunmixingofhyperspectralimages |