Deep Transfer Learning for Land Use and Land Cover Classification: A Comparative Study
Efficiently implementing remote sensing image classification with high spatial resolution imagery can provide significant value in land use and land cover (LULC) classification. The new advances in remote sensing and deep learning technologies have facilitated the extraction of spatiotemporal inform...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-12-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/21/23/8083 |
| _version_ | 1797507167509544960 |
|---|---|
| author | Raoof Naushad Tarunpreet Kaur Ebrahim Ghaderpour |
| author_facet | Raoof Naushad Tarunpreet Kaur Ebrahim Ghaderpour |
| author_sort | Raoof Naushad |
| collection | DOAJ |
| description | Efficiently implementing remote sensing image classification with high spatial resolution imagery can provide significant value in land use and land cover (LULC) classification. The new advances in remote sensing and deep learning technologies have facilitated the extraction of spatiotemporal information for LULC classification. Moreover, diverse disciplines of science, including remote sensing, have utilised tremendous improvements in image classification involving convolutional neural networks (CNNs) with transfer learning. In this study, instead of training CNNs from scratch, the transfer learning was applied to fine-tune pre-trained networks Visual Geometry Group (VGG16) and Wide Residual Networks (WRNs), by replacing the final layers with additional layers, for LULC classification using the red–green–blue version of the EuroSAT dataset. Moreover, the performance and computational time are compared and optimised with techniques such as early stopping, gradient clipping, adaptive learning rates, and data augmentation. The proposed approaches have addressed the limited-data problem, and very good accuracies were achieved. The results show that the proposed method based on WRNs outperformed the previous best results in terms of computational efficiency and accuracy, by achieving 99.17%. |
| first_indexed | 2024-03-10T04:44:47Z |
| format | Article |
| id | doaj.art-bd6bd6f1bea748fca597f0257604c140 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-10T04:44:47Z |
| publishDate | 2021-12-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-bd6bd6f1bea748fca597f0257604c1402023-11-23T03:03:44ZengMDPI AGSensors1424-82202021-12-012123808310.3390/s21238083Deep Transfer Learning for Land Use and Land Cover Classification: A Comparative StudyRaoof Naushad0Tarunpreet Kaur1Ebrahim Ghaderpour2Accubits Invent—Artificial Intelligence R&D Lab, Accubits Technologies Inc., Trivandrum 695581, IndiaDepartment of Biomedical Science, Acharya Narendra Dev College, University of Delhi, Delhi 110019, IndiaDepartment of Geomatics Engineering, University of Calgary, 2500 University Drive NW, Calgary, AB T2N 1N4, CanadaEfficiently implementing remote sensing image classification with high spatial resolution imagery can provide significant value in land use and land cover (LULC) classification. The new advances in remote sensing and deep learning technologies have facilitated the extraction of spatiotemporal information for LULC classification. Moreover, diverse disciplines of science, including remote sensing, have utilised tremendous improvements in image classification involving convolutional neural networks (CNNs) with transfer learning. In this study, instead of training CNNs from scratch, the transfer learning was applied to fine-tune pre-trained networks Visual Geometry Group (VGG16) and Wide Residual Networks (WRNs), by replacing the final layers with additional layers, for LULC classification using the red–green–blue version of the EuroSAT dataset. Moreover, the performance and computational time are compared and optimised with techniques such as early stopping, gradient clipping, adaptive learning rates, and data augmentation. The proposed approaches have addressed the limited-data problem, and very good accuracies were achieved. The results show that the proposed method based on WRNs outperformed the previous best results in terms of computational efficiency and accuracy, by achieving 99.17%.https://www.mdpi.com/1424-8220/21/23/8083land use classificationland cover classificationremote sensingsatellite imageryEuroSATearth observation |
| spellingShingle | Raoof Naushad Tarunpreet Kaur Ebrahim Ghaderpour Deep Transfer Learning for Land Use and Land Cover Classification: A Comparative Study Sensors land use classification land cover classification remote sensing satellite imagery EuroSAT earth observation |
| title | Deep Transfer Learning for Land Use and Land Cover Classification: A Comparative Study |
| title_full | Deep Transfer Learning for Land Use and Land Cover Classification: A Comparative Study |
| title_fullStr | Deep Transfer Learning for Land Use and Land Cover Classification: A Comparative Study |
| title_full_unstemmed | Deep Transfer Learning for Land Use and Land Cover Classification: A Comparative Study |
| title_short | Deep Transfer Learning for Land Use and Land Cover Classification: A Comparative Study |
| title_sort | deep transfer learning for land use and land cover classification a comparative study |
| topic | land use classification land cover classification remote sensing satellite imagery EuroSAT earth observation |
| url | https://www.mdpi.com/1424-8220/21/23/8083 |
| work_keys_str_mv | AT raoofnaushad deeptransferlearningforlanduseandlandcoverclassificationacomparativestudy AT tarunpreetkaur deeptransferlearningforlanduseandlandcoverclassificationacomparativestudy AT ebrahimghaderpour deeptransferlearningforlanduseandlandcoverclassificationacomparativestudy |