PCCN-MSS: Parallel Convolutional Classification Network Combined Multi-Spatial Scale and Spectral Features for UAV-Borne Hyperspectral With High Spatial Resolution Imagery
Hyperspectral remote sensing images with high spatial resolution (H<sup>2</sup> imagery) have an abundant spatial-spectral information, holding tremendous potential for remote sensing fine-grained monitoring and classification. However, challenges such as high spatial heterogeneity, seve...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2024-01-01
|
| Series: | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10448531/ |
| _version_ | 1797243473160568832 |
|---|---|
| author | Linhuan Jiang Zhen Zhang Bo-Hui Tang Lehao Huang Bingru Zhang |
| author_facet | Linhuan Jiang Zhen Zhang Bo-Hui Tang Lehao Huang Bingru Zhang |
| author_sort | Linhuan Jiang |
| collection | DOAJ |
| description | Hyperspectral remote sensing images with high spatial resolution (H<sup>2</sup> imagery) have an abundant spatial-spectral information, holding tremendous potential for remote sensing fine-grained monitoring and classification. However, challenges such as high spatial heterogeneity, severe intra-class spectral variability, and poor signal-to-noise ratio especially in unmanned aerial vehicle (UAV) hyperspectral imagery constrain and hinder the performance of fine-grained classification. Convolutional neural network (CNN) emerges as a formidable and excellent tool for image mining and feature extraction, offering effective utility for land cover classification. In this article, a parallel convolutional classification network model based on multimodal filters [including independent component analysis (ICA)-two-dimensional (2-D)-FPN and spectral attention (SA)-3-D-CNN branching structures] PCCN-MSS is proposed for precise H<sup>2</sup> imagery classification. The ICA-2-D-FPN branch integrates ICA into 2-D-CNN to extract the multispatial scale and spectral information of H<sup>2</sup> imagery by feature pyramid networks, meanwhile, the SA-3-D-CNN branch is designed to extract the spatial and spectral information by combining SA mechanism and 3-D-CNN. Taking hyperspectral imagery of UAVs containing vegetation and artifactual material ground as an example, the proposed PCCN-MSS model achieves an overall accuracy of 78.18%, which outperforms by 9.58% to the compared methods. The proposed PCCN-MSS method can mitigate the classification issues of severe salt-and-pepper noise and inaccurate boundary, delivering more satisfactory classification results with robust classification performance and remarkable advantages for H<sup>2</sup> imagery. |
| first_indexed | 2024-04-24T18:55:40Z |
| format | Article |
| id | doaj.art-dd43f3c17ee442789946a528686948c3 |
| institution | Directory Open Access Journal |
| issn | 2151-1535 |
| language | English |
| last_indexed | 2024-04-24T18:55:40Z |
| publishDate | 2024-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| spelling | doaj.art-dd43f3c17ee442789946a528686948c32024-03-26T17:35:15ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing2151-15352024-01-01176529654310.1109/JSTARS.2024.337063210448531PCCN-MSS: Parallel Convolutional Classification Network Combined Multi-Spatial Scale and Spectral Features for UAV-Borne Hyperspectral With High Spatial Resolution ImageryLinhuan Jiang0https://orcid.org/0009-0006-7723-478XZhen Zhang1https://orcid.org/0000-0002-2300-7112Bo-Hui Tang2https://orcid.org/0000-0002-1918-5346Lehao Huang3https://orcid.org/0009-0008-4239-2088Bingru Zhang4https://orcid.org/0009-0009-5721-1996Faculty of Land Resource Engineering, Kunming University of Science and Technology, Kunming, ChinaFaculty of Land Resource Engineering, Kunming University of Science and Technology, Kunming, ChinaFaculty of Land Resource Engineering, Kunming University of Science and Technology, Kunming, ChinaFaculty of Land Resource Engineering, Kunming University of Science and Technology, Kunming, ChinaFaculty of Land Resource Engineering, Kunming University of Science and Technology, Kunming, ChinaHyperspectral remote sensing images with high spatial resolution (H<sup>2</sup> imagery) have an abundant spatial-spectral information, holding tremendous potential for remote sensing fine-grained monitoring and classification. However, challenges such as high spatial heterogeneity, severe intra-class spectral variability, and poor signal-to-noise ratio especially in unmanned aerial vehicle (UAV) hyperspectral imagery constrain and hinder the performance of fine-grained classification. Convolutional neural network (CNN) emerges as a formidable and excellent tool for image mining and feature extraction, offering effective utility for land cover classification. In this article, a parallel convolutional classification network model based on multimodal filters [including independent component analysis (ICA)-two-dimensional (2-D)-FPN and spectral attention (SA)-3-D-CNN branching structures] PCCN-MSS is proposed for precise H<sup>2</sup> imagery classification. The ICA-2-D-FPN branch integrates ICA into 2-D-CNN to extract the multispatial scale and spectral information of H<sup>2</sup> imagery by feature pyramid networks, meanwhile, the SA-3-D-CNN branch is designed to extract the spatial and spectral information by combining SA mechanism and 3-D-CNN. Taking hyperspectral imagery of UAVs containing vegetation and artifactual material ground as an example, the proposed PCCN-MSS model achieves an overall accuracy of 78.18%, which outperforms by 9.58% to the compared methods. The proposed PCCN-MSS method can mitigate the classification issues of severe salt-and-pepper noise and inaccurate boundary, delivering more satisfactory classification results with robust classification performance and remarkable advantages for H<sup>2</sup> imagery.https://ieeexplore.ieee.org/document/10448531/Feature pyramid networks (FPNs)image classificationparallel convolutional classification networkspectral attention (SA)unmanned aerial vehicle (UAV)-borne hyperspectral imagery |
| spellingShingle | Linhuan Jiang Zhen Zhang Bo-Hui Tang Lehao Huang Bingru Zhang PCCN-MSS: Parallel Convolutional Classification Network Combined Multi-Spatial Scale and Spectral Features for UAV-Borne Hyperspectral With High Spatial Resolution Imagery IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing Feature pyramid networks (FPNs) image classification parallel convolutional classification network spectral attention (SA) unmanned aerial vehicle (UAV)-borne hyperspectral imagery |
| title | PCCN-MSS: Parallel Convolutional Classification Network Combined Multi-Spatial Scale and Spectral Features for UAV-Borne Hyperspectral With High Spatial Resolution Imagery |
| title_full | PCCN-MSS: Parallel Convolutional Classification Network Combined Multi-Spatial Scale and Spectral Features for UAV-Borne Hyperspectral With High Spatial Resolution Imagery |
| title_fullStr | PCCN-MSS: Parallel Convolutional Classification Network Combined Multi-Spatial Scale and Spectral Features for UAV-Borne Hyperspectral With High Spatial Resolution Imagery |
| title_full_unstemmed | PCCN-MSS: Parallel Convolutional Classification Network Combined Multi-Spatial Scale and Spectral Features for UAV-Borne Hyperspectral With High Spatial Resolution Imagery |
| title_short | PCCN-MSS: Parallel Convolutional Classification Network Combined Multi-Spatial Scale and Spectral Features for UAV-Borne Hyperspectral With High Spatial Resolution Imagery |
| title_sort | pccn mss parallel convolutional classification network combined multi spatial scale and spectral features for uav borne hyperspectral with high spatial resolution imagery |
| topic | Feature pyramid networks (FPNs) image classification parallel convolutional classification network spectral attention (SA) unmanned aerial vehicle (UAV)-borne hyperspectral imagery |
| url | https://ieeexplore.ieee.org/document/10448531/ |
| work_keys_str_mv | AT linhuanjiang pccnmssparallelconvolutionalclassificationnetworkcombinedmultispatialscaleandspectralfeaturesforuavbornehyperspectralwithhighspatialresolutionimagery AT zhenzhang pccnmssparallelconvolutionalclassificationnetworkcombinedmultispatialscaleandspectralfeaturesforuavbornehyperspectralwithhighspatialresolutionimagery AT bohuitang pccnmssparallelconvolutionalclassificationnetworkcombinedmultispatialscaleandspectralfeaturesforuavbornehyperspectralwithhighspatialresolutionimagery AT lehaohuang pccnmssparallelconvolutionalclassificationnetworkcombinedmultispatialscaleandspectralfeaturesforuavbornehyperspectralwithhighspatialresolutionimagery AT bingruzhang pccnmssparallelconvolutionalclassificationnetworkcombinedmultispatialscaleandspectralfeaturesforuavbornehyperspectralwithhighspatialresolutionimagery |