Identification of Earth’s surface objects using ensembles of convolutional neural networks.
The paper proposes an identification technique of objects on the Earth’s surface images based on combination of machine learning methods. Different variants of multi-layer convolutional neural networks and support vector machines are considered as original models. A hybrid convolutional neural netwo...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | Belarusian |
| Published: |
Belarusian State University
2021-08-01
|
| Series: | Журнал Белорусского государственного университета: Математика, информатика |
| Subjects: | |
| Online Access: | https://journals.bsu.by/index.php/mathematics/article/view/3703 |
| _version_ | 1818385274288209920 |
|---|---|
| author | Evgenii E. Marushko Alexander A. Doudkin Xiangtao Zheng |
| author_facet | Evgenii E. Marushko Alexander A. Doudkin Xiangtao Zheng |
| author_sort | Evgenii E. Marushko |
| collection | DOAJ |
| description | The paper proposes an identification technique of objects on the Earth’s surface images based on combination of machine learning methods. Different variants of multi-layer convolutional neural networks and support vector machines are considered as original models. A hybrid convolutional neural network that combines features extracted by the neural network and experts is proposed. Optimal values of hyperparameters of the models are calculated by grid search methods using k-fold cross-validation. The possibility of improving the accuracy of identification based on the ensembles of these models is shown. Effectiveness of the proposed technique is demonstrated by the example of images obtained by synthetic aperture radar. |
| first_indexed | 2024-12-14T03:35:33Z |
| format | Article |
| id | doaj.art-ceecbf199742474b8b556e47a6729f05 |
| institution | Directory Open Access Journal |
| issn | 2520-6508 2617-3956 |
| language | Belarusian |
| last_indexed | 2024-12-14T03:35:33Z |
| publishDate | 2021-08-01 |
| publisher | Belarusian State University |
| record_format | Article |
| series | Журнал Белорусского государственного университета: Математика, информатика |
| spelling | doaj.art-ceecbf199742474b8b556e47a6729f052022-12-21T23:18:38ZbelBelarusian State UniversityЖурнал Белорусского государственного университета: Математика, информатика2520-65082617-39562021-08-01211412310.33581/2520-6508-2021-2-114-1233703Identification of Earth’s surface objects using ensembles of convolutional neural networks.Evgenii E. Marushko0Alexander A. Doudkin1Xiangtao Zheng2United Institute of Informatics Problems, National Academy of Sciences of Belarus, 6 Surhanava Street, Minsk 220012, BelarusUnited Institute of Informatics Problems, National Academy of Sciences of Belarus, 6 Surhanava Street, Minsk 220012, Belarus Xiʼan Institute of Optics and Precision Mechanics, Chinese Academy of Sciences, Shaanxi, Xiʼan 710119, ChinaThe paper proposes an identification technique of objects on the Earth’s surface images based on combination of machine learning methods. Different variants of multi-layer convolutional neural networks and support vector machines are considered as original models. A hybrid convolutional neural network that combines features extracted by the neural network and experts is proposed. Optimal values of hyperparameters of the models are calculated by grid search methods using k-fold cross-validation. The possibility of improving the accuracy of identification based on the ensembles of these models is shown. Effectiveness of the proposed technique is demonstrated by the example of images obtained by synthetic aperture radar.https://journals.bsu.by/index.php/mathematics/article/view/3703convolutional neural networksupport vector machineneural network ensembleearth’s surface imageremote sensingidentificationsynthetic aperture radar |
| spellingShingle | Evgenii E. Marushko Alexander A. Doudkin Xiangtao Zheng Identification of Earth’s surface objects using ensembles of convolutional neural networks. Журнал Белорусского государственного университета: Математика, информатика convolutional neural network support vector machine neural network ensemble earth’s surface image remote sensing identification synthetic aperture radar |
| title | Identification of Earth’s surface objects using ensembles of convolutional neural networks. |
| title_full | Identification of Earth’s surface objects using ensembles of convolutional neural networks. |
| title_fullStr | Identification of Earth’s surface objects using ensembles of convolutional neural networks. |
| title_full_unstemmed | Identification of Earth’s surface objects using ensembles of convolutional neural networks. |
| title_short | Identification of Earth’s surface objects using ensembles of convolutional neural networks. |
| title_sort | identification of earth s surface objects using ensembles of convolutional neural networks |
| topic | convolutional neural network support vector machine neural network ensemble earth’s surface image remote sensing identification synthetic aperture radar |
| url | https://journals.bsu.by/index.php/mathematics/article/view/3703 |
| work_keys_str_mv | AT evgeniiemarushko identificationofearthssurfaceobjectsusingensemblesofconvolutionalneuralnetworks AT alexanderadoudkin identificationofearthssurfaceobjectsusingensemblesofconvolutionalneuralnetworks AT xiangtaozheng identificationofearthssurfaceobjectsusingensemblesofconvolutionalneuralnetworks |