A new band selection framework for hyperspectral remote sensing image classification
Abstract Dimensionality Reduction (DR) is an indispensable step to enhance classifier accuracy with data redundancy in hyperspectral images (HSI). This paper proposes a framework for DR that combines band selection (BS) and effective spatial features. The conventional clustering methods for BS typic...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2024-12-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-024-83118-8 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841559513828687872 |
|---|---|
| author | B. L. N. Phaneendra Kumar Radhesyam Vaddi Prabukumar Manoharan L. Agilandeeswari V. Sangeetha |
| author_facet | B. L. N. Phaneendra Kumar Radhesyam Vaddi Prabukumar Manoharan L. Agilandeeswari V. Sangeetha |
| author_sort | B. L. N. Phaneendra Kumar |
| collection | DOAJ |
| description | Abstract Dimensionality Reduction (DR) is an indispensable step to enhance classifier accuracy with data redundancy in hyperspectral images (HSI). This paper proposes a framework for DR that combines band selection (BS) and effective spatial features. The conventional clustering methods for BS typically face hard encounters when we have a less data items matched to the dimensionality of the accompanying feature space. So, to fully mine the effective information, BS is established using dual partitioning and ranking. The bands from the dual partitioning have undergone informative band selection via ranking. The reduced band subset is then given to a hemispherical reflectance-based spatial filter. Then, finally, a Convolutional Neural Network (CNN) is used for effective classification by incorporating three-dimensional convolutions. On a set of three hyperspectral datasets - Indian Pines, Salinas, and KSC, the proposed method was tested with different state-of-the-art techniques. The classification results are compared using quantitative and qualitative measures. The reported overall accuracy is 99.92% on Indian Pines, 99.94% on Salinas, and 97.23% on the KSC dataset. Also, the Mean Spectral Divergence values are 42.4, 63.75, and 41.2 on the three datasets respectively, which signifies the effectiveness of band selection. The results have clearly shown the impact of the band selection proposed and can be utilized for a wide variety of applications. |
| format | Article |
| id | doaj-art-b6951cc115bb4b73a9d6abb4149ee936 |
| institution | Kabale University |
| issn | 2045-2322 |
| language | English |
| publishDate | 2024-12-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj-art-b6951cc115bb4b73a9d6abb4149ee9362025-01-05T12:26:47ZengNature PortfolioScientific Reports2045-23222024-12-0114112110.1038/s41598-024-83118-8A new band selection framework for hyperspectral remote sensing image classificationB. L. N. Phaneendra Kumar0Radhesyam Vaddi1Prabukumar Manoharan2L. Agilandeeswari3V. Sangeetha4Department of Computer Science and Engineering, Koneru Lakshmaiah Education FoundationDepartment of Information Technology, Velagapudi Ramakrishna Siddhartha Engineering College (Deemed to be University)School of Computer Science Engineering and Information Systems, Vellore Institute of TechnologySchool of Computer Science Engineering and Information Systems, Vellore Institute of TechnologySchool of Computer Science Engineering and Information Systems, Vellore Institute of TechnologyAbstract Dimensionality Reduction (DR) is an indispensable step to enhance classifier accuracy with data redundancy in hyperspectral images (HSI). This paper proposes a framework for DR that combines band selection (BS) and effective spatial features. The conventional clustering methods for BS typically face hard encounters when we have a less data items matched to the dimensionality of the accompanying feature space. So, to fully mine the effective information, BS is established using dual partitioning and ranking. The bands from the dual partitioning have undergone informative band selection via ranking. The reduced band subset is then given to a hemispherical reflectance-based spatial filter. Then, finally, a Convolutional Neural Network (CNN) is used for effective classification by incorporating three-dimensional convolutions. On a set of three hyperspectral datasets - Indian Pines, Salinas, and KSC, the proposed method was tested with different state-of-the-art techniques. The classification results are compared using quantitative and qualitative measures. The reported overall accuracy is 99.92% on Indian Pines, 99.94% on Salinas, and 97.23% on the KSC dataset. Also, the Mean Spectral Divergence values are 42.4, 63.75, and 41.2 on the three datasets respectively, which signifies the effectiveness of band selection. The results have clearly shown the impact of the band selection proposed and can be utilized for a wide variety of applications.https://doi.org/10.1038/s41598-024-83118-8HyperspectralPCANonlinearSmoothing filter-weighted least squaresHyperbolic sigmoidSpectral |
| spellingShingle | B. L. N. Phaneendra Kumar Radhesyam Vaddi Prabukumar Manoharan L. Agilandeeswari V. Sangeetha A new band selection framework for hyperspectral remote sensing image classification Scientific Reports Hyperspectral PCA Nonlinear Smoothing filter-weighted least squares Hyperbolic sigmoid Spectral |
| title | A new band selection framework for hyperspectral remote sensing image classification |
| title_full | A new band selection framework for hyperspectral remote sensing image classification |
| title_fullStr | A new band selection framework for hyperspectral remote sensing image classification |
| title_full_unstemmed | A new band selection framework for hyperspectral remote sensing image classification |
| title_short | A new band selection framework for hyperspectral remote sensing image classification |
| title_sort | new band selection framework for hyperspectral remote sensing image classification |
| topic | Hyperspectral PCA Nonlinear Smoothing filter-weighted least squares Hyperbolic sigmoid Spectral |
| url | https://doi.org/10.1038/s41598-024-83118-8 |
| work_keys_str_mv | AT blnphaneendrakumar anewbandselectionframeworkforhyperspectralremotesensingimageclassification AT radhesyamvaddi anewbandselectionframeworkforhyperspectralremotesensingimageclassification AT prabukumarmanoharan anewbandselectionframeworkforhyperspectralremotesensingimageclassification AT lagilandeeswari anewbandselectionframeworkforhyperspectralremotesensingimageclassification AT vsangeetha anewbandselectionframeworkforhyperspectralremotesensingimageclassification AT blnphaneendrakumar newbandselectionframeworkforhyperspectralremotesensingimageclassification AT radhesyamvaddi newbandselectionframeworkforhyperspectralremotesensingimageclassification AT prabukumarmanoharan newbandselectionframeworkforhyperspectralremotesensingimageclassification AT lagilandeeswari newbandselectionframeworkforhyperspectralremotesensingimageclassification AT vsangeetha newbandselectionframeworkforhyperspectralremotesensingimageclassification |