Tree Species Classification from Airborne Hyperspectral Images Using Spatial–Spectral Network
Tree species identification is a critical component of forest resource monitoring, and timely and accurate acquisition of tree species information is the basis for sustainable forest management and resource assessment. Airborne hyperspectral images have rich spectral and spatial information and can...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-12-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/15/24/5679 |
| _version_ | 1797379503771615232 |
|---|---|
| author | Chengchao Hou Zhengjun Liu Yiming Chen Shuo Wang Aixia Liu |
| author_facet | Chengchao Hou Zhengjun Liu Yiming Chen Shuo Wang Aixia Liu |
| author_sort | Chengchao Hou |
| collection | DOAJ |
| description | Tree species identification is a critical component of forest resource monitoring, and timely and accurate acquisition of tree species information is the basis for sustainable forest management and resource assessment. Airborne hyperspectral images have rich spectral and spatial information and can detect subtle differences among tree species. To fully utilize the advantages of hyperspectral images, we propose a double-branch spatial–spectral joint network based on the SimAM attention mechanism for tree species classification. This method achieved high classification accuracy on three tree species datasets (93.31% OA value obtained in the TEF dataset, 95.7% in the Tiegang Reservoir dataset, and 98.82% in the Xiongan New Area dataset). The network consists of three parts: spectral branch, spatial branch, and feature fusion, and both branches make full use of the spatial–spectral information of pixels to avoid the loss of information. In addition, the SimAM attention mechanism is added to the feature fusion part of the network to refine the features to extract more critical features for high-precision tree species classification. To validate the robustness of the proposed method, we compared this method with other advanced classification methods through a series of experiments. The results show that: (1) Compared with traditional machine learning methods (SVM, RF) and other state-of-the-art deep learning methods, the proposed method achieved the highest classification accuracy in all three tree datasets. (2) Combining spatial and spectral information and incorporating the SimAM attention mechanism into the network can improve the classification accuracy of tree species, and the classification performance of the double-branch network is better than that of the single-branch network. (3) The proposed method obtains the highest accuracy under different training sample proportions, and does not change significantly with different training sample proportions, which are stable. This study demonstrates that high-precision tree species classification can be achieved using airborne hyperspectral images and the methods proposed in this study, which have great potential in investigating and monitoring forest resources. |
| first_indexed | 2024-03-08T20:24:13Z |
| format | Article |
| id | doaj.art-897da6d410b6482c9fec24b6283e803e |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-08T20:24:13Z |
| publishDate | 2023-12-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-897da6d410b6482c9fec24b6283e803e2023-12-22T14:38:58ZengMDPI AGRemote Sensing2072-42922023-12-011524567910.3390/rs15245679Tree Species Classification from Airborne Hyperspectral Images Using Spatial–Spectral NetworkChengchao Hou0Zhengjun Liu1Yiming Chen2Shuo Wang3Aixia Liu4Institute of Photogrammetry and Remote Sensing, Chinese Academy of Surveying and Mapping, Beijing 100036, ChinaInstitute of Photogrammetry and Remote Sensing, Chinese Academy of Surveying and Mapping, Beijing 100036, ChinaInstitute of Photogrammetry and Remote Sensing, Chinese Academy of Surveying and Mapping, Beijing 100036, ChinaInstitute of Photogrammetry and Remote Sensing, Chinese Academy of Surveying and Mapping, Beijing 100036, ChinaLand Satellite Remote Sensing Application Center, Ministry of Natural Resources, Beijing 100048, ChinaTree species identification is a critical component of forest resource monitoring, and timely and accurate acquisition of tree species information is the basis for sustainable forest management and resource assessment. Airborne hyperspectral images have rich spectral and spatial information and can detect subtle differences among tree species. To fully utilize the advantages of hyperspectral images, we propose a double-branch spatial–spectral joint network based on the SimAM attention mechanism for tree species classification. This method achieved high classification accuracy on three tree species datasets (93.31% OA value obtained in the TEF dataset, 95.7% in the Tiegang Reservoir dataset, and 98.82% in the Xiongan New Area dataset). The network consists of three parts: spectral branch, spatial branch, and feature fusion, and both branches make full use of the spatial–spectral information of pixels to avoid the loss of information. In addition, the SimAM attention mechanism is added to the feature fusion part of the network to refine the features to extract more critical features for high-precision tree species classification. To validate the robustness of the proposed method, we compared this method with other advanced classification methods through a series of experiments. The results show that: (1) Compared with traditional machine learning methods (SVM, RF) and other state-of-the-art deep learning methods, the proposed method achieved the highest classification accuracy in all three tree datasets. (2) Combining spatial and spectral information and incorporating the SimAM attention mechanism into the network can improve the classification accuracy of tree species, and the classification performance of the double-branch network is better than that of the single-branch network. (3) The proposed method obtains the highest accuracy under different training sample proportions, and does not change significantly with different training sample proportions, which are stable. This study demonstrates that high-precision tree species classification can be achieved using airborne hyperspectral images and the methods proposed in this study, which have great potential in investigating and monitoring forest resources.https://www.mdpi.com/2072-4292/15/24/5679tree species classificationhyperspectral imagesdeep learningspatial–spectral informationattention mechanism |
| spellingShingle | Chengchao Hou Zhengjun Liu Yiming Chen Shuo Wang Aixia Liu Tree Species Classification from Airborne Hyperspectral Images Using Spatial–Spectral Network Remote Sensing tree species classification hyperspectral images deep learning spatial–spectral information attention mechanism |
| title | Tree Species Classification from Airborne Hyperspectral Images Using Spatial–Spectral Network |
| title_full | Tree Species Classification from Airborne Hyperspectral Images Using Spatial–Spectral Network |
| title_fullStr | Tree Species Classification from Airborne Hyperspectral Images Using Spatial–Spectral Network |
| title_full_unstemmed | Tree Species Classification from Airborne Hyperspectral Images Using Spatial–Spectral Network |
| title_short | Tree Species Classification from Airborne Hyperspectral Images Using Spatial–Spectral Network |
| title_sort | tree species classification from airborne hyperspectral images using spatial spectral network |
| topic | tree species classification hyperspectral images deep learning spatial–spectral information attention mechanism |
| url | https://www.mdpi.com/2072-4292/15/24/5679 |
| work_keys_str_mv | AT chengchaohou treespeciesclassificationfromairbornehyperspectralimagesusingspatialspectralnetwork AT zhengjunliu treespeciesclassificationfromairbornehyperspectralimagesusingspatialspectralnetwork AT yimingchen treespeciesclassificationfromairbornehyperspectralimagesusingspatialspectralnetwork AT shuowang treespeciesclassificationfromairbornehyperspectralimagesusingspatialspectralnetwork AT aixialiu treespeciesclassificationfromairbornehyperspectralimagesusingspatialspectralnetwork |