Multi-Source Attention U-Net: A Novel Deep Learning Framework for the Land Use and Soil Salinization Classification of Keriya Oasis in China with RADARSAT-2 and Landsat-8 Data
Soil salinization significantly impacts global agricultural productivity, contributing to desertification and land degradation; thus, rapid regional monitoring of soil salinization is crucial for agricultural production and sustainable management. With advancements in artificial intelligence, the ef...
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-03-01
|
| Series: | Land |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2073-445X/14/3/649 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850205284858855424 |
|---|---|
| author | Yang Xiang Ilyas Nurmemet Xiaobo Lv Xinru Yu Aoxiang Gu Aihepa Aihaiti Shiqin Li |
| author_facet | Yang Xiang Ilyas Nurmemet Xiaobo Lv Xinru Yu Aoxiang Gu Aihepa Aihaiti Shiqin Li |
| author_sort | Yang Xiang |
| collection | DOAJ |
| description | Soil salinization significantly impacts global agricultural productivity, contributing to desertification and land degradation; thus, rapid regional monitoring of soil salinization is crucial for agricultural production and sustainable management. With advancements in artificial intelligence, the efficiency and precision of deep learning classification models applied to remote sensing imagery have been demonstrated. Given the limited feature learning capability of traditional machine learning, this study introduces an innovative deep fusion U-Net model called MSA-U-Net (Multi-Source Attention U-Net) incorporating a Convolutional Block Attention Module (CBAM) within the skip connections to improve feature extraction and fusion. A salinized soil classification dataset was developed by combining spectral indices obtained from Landsat-8 Operational Land Imager (OLI) data and polarimetric scattering features extracted from RADARSAT-2 data using polarization target decomposition. To select optimal features, the Boruta algorithm was employed to rank features, selecting the top eight features to construct a multispectral (MS) dataset, a synthetic aperture radar (SAR) dataset, and an MS + SAR dataset. Furthermore, Support Vector Machine (SVM), Random Forest (RF), K-Nearest Neighbor (KNN), and deep learning methods including U-Net and MSA-U-Net were employed to identify the different degrees of salinized soil. The results indicated that the MS + SAR dataset outperformed the MS dataset, with the inclusion of the SAR band resulting in an Overall Accuracy (OA) increase of 1.94–7.77%. Moreover, the MS + SAR MSA-U-Net, in comparison to traditional machine learning methods and the baseline model, improved the OA and Kappa coefficient by 8.24% to 12.55% and 0.08 to 0.15, respectively. The results demonstrate that the MSA-U-Net outperformed traditional models, indicating the potential of integrating multi-source data with deep learning techniques for monitoring soil salinity. |
| format | Article |
| id | doaj-art-4b6e4ab0263f405388f91039cf58b51d |
| institution | OA Journals |
| issn | 2073-445X |
| language | English |
| publishDate | 2025-03-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Land |
| spelling | doaj-art-4b6e4ab0263f405388f91039cf58b51d2025-08-20T02:11:08ZengMDPI AGLand2073-445X2025-03-0114364910.3390/land14030649Multi-Source Attention U-Net: A Novel Deep Learning Framework for the Land Use and Soil Salinization Classification of Keriya Oasis in China with RADARSAT-2 and Landsat-8 DataYang Xiang0Ilyas Nurmemet1Xiaobo Lv2Xinru Yu3Aoxiang Gu4Aihepa Aihaiti5Shiqin Li6College of Geography and Remote Sensing Sciences, Xinjiang University, Urumqi 830046, ChinaCollege of Geography and Remote Sensing Sciences, Xinjiang University, Urumqi 830046, ChinaCollege of Geography and Remote Sensing Sciences, Xinjiang University, Urumqi 830046, ChinaCollege of Geography and Remote Sensing Sciences, Xinjiang University, Urumqi 830046, ChinaCollege of Geography and Remote Sensing Sciences, Xinjiang University, Urumqi 830046, ChinaCollege of Geography and Remote Sensing Sciences, Xinjiang University, Urumqi 830046, ChinaCollege of Geography and Remote Sensing Sciences, Xinjiang University, Urumqi 830046, ChinaSoil salinization significantly impacts global agricultural productivity, contributing to desertification and land degradation; thus, rapid regional monitoring of soil salinization is crucial for agricultural production and sustainable management. With advancements in artificial intelligence, the efficiency and precision of deep learning classification models applied to remote sensing imagery have been demonstrated. Given the limited feature learning capability of traditional machine learning, this study introduces an innovative deep fusion U-Net model called MSA-U-Net (Multi-Source Attention U-Net) incorporating a Convolutional Block Attention Module (CBAM) within the skip connections to improve feature extraction and fusion. A salinized soil classification dataset was developed by combining spectral indices obtained from Landsat-8 Operational Land Imager (OLI) data and polarimetric scattering features extracted from RADARSAT-2 data using polarization target decomposition. To select optimal features, the Boruta algorithm was employed to rank features, selecting the top eight features to construct a multispectral (MS) dataset, a synthetic aperture radar (SAR) dataset, and an MS + SAR dataset. Furthermore, Support Vector Machine (SVM), Random Forest (RF), K-Nearest Neighbor (KNN), and deep learning methods including U-Net and MSA-U-Net were employed to identify the different degrees of salinized soil. The results indicated that the MS + SAR dataset outperformed the MS dataset, with the inclusion of the SAR band resulting in an Overall Accuracy (OA) increase of 1.94–7.77%. Moreover, the MS + SAR MSA-U-Net, in comparison to traditional machine learning methods and the baseline model, improved the OA and Kappa coefficient by 8.24% to 12.55% and 0.08 to 0.15, respectively. The results demonstrate that the MSA-U-Net outperformed traditional models, indicating the potential of integrating multi-source data with deep learning techniques for monitoring soil salinity.https://www.mdpi.com/2073-445X/14/3/649soil salinityRADARSAT-2classificationpolarimetric decompositiondeep learning |
| spellingShingle | Yang Xiang Ilyas Nurmemet Xiaobo Lv Xinru Yu Aoxiang Gu Aihepa Aihaiti Shiqin Li Multi-Source Attention U-Net: A Novel Deep Learning Framework for the Land Use and Soil Salinization Classification of Keriya Oasis in China with RADARSAT-2 and Landsat-8 Data Land soil salinity RADARSAT-2 classification polarimetric decomposition deep learning |
| title | Multi-Source Attention U-Net: A Novel Deep Learning Framework for the Land Use and Soil Salinization Classification of Keriya Oasis in China with RADARSAT-2 and Landsat-8 Data |
| title_full | Multi-Source Attention U-Net: A Novel Deep Learning Framework for the Land Use and Soil Salinization Classification of Keriya Oasis in China with RADARSAT-2 and Landsat-8 Data |
| title_fullStr | Multi-Source Attention U-Net: A Novel Deep Learning Framework for the Land Use and Soil Salinization Classification of Keriya Oasis in China with RADARSAT-2 and Landsat-8 Data |
| title_full_unstemmed | Multi-Source Attention U-Net: A Novel Deep Learning Framework for the Land Use and Soil Salinization Classification of Keriya Oasis in China with RADARSAT-2 and Landsat-8 Data |
| title_short | Multi-Source Attention U-Net: A Novel Deep Learning Framework for the Land Use and Soil Salinization Classification of Keriya Oasis in China with RADARSAT-2 and Landsat-8 Data |
| title_sort | multi source attention u net a novel deep learning framework for the land use and soil salinization classification of keriya oasis in china with radarsat 2 and landsat 8 data |
| topic | soil salinity RADARSAT-2 classification polarimetric decomposition deep learning |
| url | https://www.mdpi.com/2073-445X/14/3/649 |
| work_keys_str_mv | AT yangxiang multisourceattentionunetanoveldeeplearningframeworkforthelanduseandsoilsalinizationclassificationofkeriyaoasisinchinawithradarsat2andlandsat8data AT ilyasnurmemet multisourceattentionunetanoveldeeplearningframeworkforthelanduseandsoilsalinizationclassificationofkeriyaoasisinchinawithradarsat2andlandsat8data AT xiaobolv multisourceattentionunetanoveldeeplearningframeworkforthelanduseandsoilsalinizationclassificationofkeriyaoasisinchinawithradarsat2andlandsat8data AT xinruyu multisourceattentionunetanoveldeeplearningframeworkforthelanduseandsoilsalinizationclassificationofkeriyaoasisinchinawithradarsat2andlandsat8data AT aoxianggu multisourceattentionunetanoveldeeplearningframeworkforthelanduseandsoilsalinizationclassificationofkeriyaoasisinchinawithradarsat2andlandsat8data AT aihepaaihaiti multisourceattentionunetanoveldeeplearningframeworkforthelanduseandsoilsalinizationclassificationofkeriyaoasisinchinawithradarsat2andlandsat8data AT shiqinli multisourceattentionunetanoveldeeplearningframeworkforthelanduseandsoilsalinizationclassificationofkeriyaoasisinchinawithradarsat2andlandsat8data |