Deep Transfer Learning of Satellite Imagery for Land Use and Land Cover Classification
Deep learning has been instrumental in solving difficult problems by automatically learning, from sample data, the rules (algorithms) that map an input to its respective output. Purpose: Perform land use landcover (LULC) classification using the training data of satellite imagery for Moscow region a...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Russian Academy of Sciences, St. Petersburg Federal Research Center
2022-09-01
|
| Series: | Информатика и автоматизация |
| Subjects: | |
| Online Access: | http://ia.spcras.ru/index.php/sp/article/view/15395 |
| _version_ | 1797699692302172160 |
|---|---|
| author | Teklay Yifter Yury Razoumny Vasiliy Lobanov |
| author_facet | Teklay Yifter Yury Razoumny Vasiliy Lobanov |
| author_sort | Teklay Yifter |
| collection | DOAJ |
| description | Deep learning has been instrumental in solving difficult problems by automatically learning, from sample data, the rules (algorithms) that map an input to its respective output. Purpose: Perform land use landcover (LULC) classification using the training data of satellite imagery for Moscow region and compare the accuracy attained from different models. Methods: The accuracy attained for LULC classification using deep learning algorithm and satellite imagery data is dependent on both the model and the training dataset used. We have used state-of-the-art deep learning models and transfer learning, together with dataset appropriate for the models. Different methods were applied to fine tuning the models with different parameters and preparing the right dataset for training, including using data augmentation. Results: Four models of deep learning from Residual Network (ResNet) and Visual Geometry Group (VGG) namely: ResNet50, ResNet152, VGG16 and VGG19 has been used with transfer learning. Further training of the models is performed with training data collected from Sentinel-2 for the Moscow region and it is found that ResNet50 has given the highest accuracy for LULC classification for this region. Practical relevance: We have developed code that train the 4 models and make classification of the input image patches into one of the 10 classes (Annual Crop, Forest, Herbaceous Vegetation, Highway, Industrial, Pasture, Permanent Crop, Residential, River, and Sea&Lake). |
| first_indexed | 2024-03-12T04:12:33Z |
| format | Article |
| id | doaj.art-007fd1eec0b74c52bbc6d0ab1680f21e |
| institution | Directory Open Access Journal |
| issn | 2713-3192 2713-3206 |
| language | English |
| last_indexed | 2024-03-12T04:12:33Z |
| publishDate | 2022-09-01 |
| publisher | Russian Academy of Sciences, St. Petersburg Federal Research Center |
| record_format | Article |
| series | Информатика и автоматизация |
| spelling | doaj.art-007fd1eec0b74c52bbc6d0ab1680f21e2023-09-03T10:52:56ZengRussian Academy of Sciences, St. Petersburg Federal Research CenterИнформатика и автоматизация2713-31922713-32062022-09-0121596398210.15622/ia.21.5.515395Deep Transfer Learning of Satellite Imagery for Land Use and Land Cover ClassificationTeklay Yifter0Yury Razoumny1Vasiliy Lobanov2Peoples' Friendship University of RussiaPeoples' Friendship University of RussiaPeoples' Friendship University of RussiaDeep learning has been instrumental in solving difficult problems by automatically learning, from sample data, the rules (algorithms) that map an input to its respective output. Purpose: Perform land use landcover (LULC) classification using the training data of satellite imagery for Moscow region and compare the accuracy attained from different models. Methods: The accuracy attained for LULC classification using deep learning algorithm and satellite imagery data is dependent on both the model and the training dataset used. We have used state-of-the-art deep learning models and transfer learning, together with dataset appropriate for the models. Different methods were applied to fine tuning the models with different parameters and preparing the right dataset for training, including using data augmentation. Results: Four models of deep learning from Residual Network (ResNet) and Visual Geometry Group (VGG) namely: ResNet50, ResNet152, VGG16 and VGG19 has been used with transfer learning. Further training of the models is performed with training data collected from Sentinel-2 for the Moscow region and it is found that ResNet50 has given the highest accuracy for LULC classification for this region. Practical relevance: We have developed code that train the 4 models and make classification of the input image patches into one of the 10 classes (Annual Crop, Forest, Herbaceous Vegetation, Highway, Industrial, Pasture, Permanent Crop, Residential, River, and Sea&Lake).http://ia.spcras.ru/index.php/sp/article/view/15395neural networksdeep transfer learningland use land cover classificationsatellite imagery |
| spellingShingle | Teklay Yifter Yury Razoumny Vasiliy Lobanov Deep Transfer Learning of Satellite Imagery for Land Use and Land Cover Classification Информатика и автоматизация neural networks deep transfer learning land use land cover classification satellite imagery |
| title | Deep Transfer Learning of Satellite Imagery for Land Use and Land Cover Classification |
| title_full | Deep Transfer Learning of Satellite Imagery for Land Use and Land Cover Classification |
| title_fullStr | Deep Transfer Learning of Satellite Imagery for Land Use and Land Cover Classification |
| title_full_unstemmed | Deep Transfer Learning of Satellite Imagery for Land Use and Land Cover Classification |
| title_short | Deep Transfer Learning of Satellite Imagery for Land Use and Land Cover Classification |
| title_sort | deep transfer learning of satellite imagery for land use and land cover classification |
| topic | neural networks deep transfer learning land use land cover classification satellite imagery |
| url | http://ia.spcras.ru/index.php/sp/article/view/15395 |
| work_keys_str_mv | AT teklayyifter deeptransferlearningofsatelliteimageryforlanduseandlandcoverclassification AT yuryrazoumny deeptransferlearningofsatelliteimageryforlanduseandlandcoverclassification AT vasiliylobanov deeptransferlearningofsatelliteimageryforlanduseandlandcoverclassification |