Gradient Boosting Machine and Object-Based CNN for Land Cover Classification
In regular convolutional neural networks (CNN), fully-connected layers act as classifiers to estimate the probabilities for each instance in classification tasks. The accuracy of CNNs can be improved by replacing fully connected layers with gradient boosting algorithms. In this regard, this study in...
| Main Authors: | , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-07-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/13/14/2709 |
| _version_ | 1797526126539571200 |
|---|---|
| author | Quang-Thanh Bui Tien-Yin Chou Thanh-Van Hoang Yao-Min Fang Ching-Yun Mu Pi-Hui Huang Vu-Dong Pham Quoc-Huy Nguyen Do Thi Ngoc Anh Van-Manh Pham Michael E. Meadows |
| author_facet | Quang-Thanh Bui Tien-Yin Chou Thanh-Van Hoang Yao-Min Fang Ching-Yun Mu Pi-Hui Huang Vu-Dong Pham Quoc-Huy Nguyen Do Thi Ngoc Anh Van-Manh Pham Michael E. Meadows |
| author_sort | Quang-Thanh Bui |
| collection | DOAJ |
| description | In regular convolutional neural networks (CNN), fully-connected layers act as classifiers to estimate the probabilities for each instance in classification tasks. The accuracy of CNNs can be improved by replacing fully connected layers with gradient boosting algorithms. In this regard, this study investigates three robust classifiers, namely XGBoost, LightGBM, and Catboost, in combination with a CNN for a land cover study in Hanoi, Vietnam. The experiments were implemented using SPOT7 imagery through (1) image segmentation and extraction of features, including spectral information and spatial metrics, (2) normalization of attribute values and generation of graphs, and (3) using graphs as the input dataset to the investigated models for classifying six land cover classes, namely House, Bare land, Vegetation, Water, Impervious Surface, and Shadow. The results show that CNN-based XGBoost (Overall accuracy = 0.8905), LightGBM (0.8956), and CatBoost (0.8956) outperform the other methods used for comparison. It can be seen that the combination of object-based image analysis and CNN-based gradient boosting algorithms significantly improves classification accuracies and can be considered as alternative methods for land cover analysis. |
| first_indexed | 2024-03-10T09:25:45Z |
| format | Article |
| id | doaj.art-7faeac37764c444497b3b0064a14383a |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-10T09:25:45Z |
| publishDate | 2021-07-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-7faeac37764c444497b3b0064a14383a2023-11-22T04:51:14ZengMDPI AGRemote Sensing2072-42922021-07-011314270910.3390/rs13142709Gradient Boosting Machine and Object-Based CNN for Land Cover ClassificationQuang-Thanh Bui0Tien-Yin Chou1Thanh-Van Hoang2Yao-Min Fang3Ching-Yun Mu4Pi-Hui Huang5Vu-Dong Pham6Quoc-Huy Nguyen7Do Thi Ngoc Anh8Van-Manh Pham9Michael E. Meadows10Center for Applied Research in Remote Sensing and GIS (CARGIS), Faculty of Geography, VNU University of Science, Vietnam National University, Hanoi, 334 Nguyen Trai, Hanoi 11416, VietnamGeographic Information Systems Research Center, Feng Chia University, Taichung 40724, TaiwanGeographic Information Systems Research Center, Feng Chia University, Taichung 40724, TaiwanGeographic Information Systems Research Center, Feng Chia University, Taichung 40724, TaiwanGeographic Information Systems Research Center, Feng Chia University, Taichung 40724, TaiwanGeographic Information Systems Research Center, Feng Chia University, Taichung 40724, TaiwanCenter for Applied Research in Remote Sensing and GIS (CARGIS), Faculty of Geography, VNU University of Science, Vietnam National University, Hanoi, 334 Nguyen Trai, Hanoi 11416, VietnamCenter for Applied Research in Remote Sensing and GIS (CARGIS), Faculty of Geography, VNU University of Science, Vietnam National University, Hanoi, 334 Nguyen Trai, Hanoi 11416, VietnamFaculty of Geography, VNU University of Science, 334 Nguyen Trai, Hanoi 11416, VietnamFaculty of Geography, VNU University of Science, 334 Nguyen Trai, Hanoi 11416, VietnamDepartment of Environmental & Geographical Science, University of Capetown, Rondebosh 7701, South AfricaIn regular convolutional neural networks (CNN), fully-connected layers act as classifiers to estimate the probabilities for each instance in classification tasks. The accuracy of CNNs can be improved by replacing fully connected layers with gradient boosting algorithms. In this regard, this study investigates three robust classifiers, namely XGBoost, LightGBM, and Catboost, in combination with a CNN for a land cover study in Hanoi, Vietnam. The experiments were implemented using SPOT7 imagery through (1) image segmentation and extraction of features, including spectral information and spatial metrics, (2) normalization of attribute values and generation of graphs, and (3) using graphs as the input dataset to the investigated models for classifying six land cover classes, namely House, Bare land, Vegetation, Water, Impervious Surface, and Shadow. The results show that CNN-based XGBoost (Overall accuracy = 0.8905), LightGBM (0.8956), and CatBoost (0.8956) outperform the other methods used for comparison. It can be seen that the combination of object-based image analysis and CNN-based gradient boosting algorithms significantly improves classification accuracies and can be considered as alternative methods for land cover analysis.https://www.mdpi.com/2072-4292/13/14/2709object-based image analysisgradient boostingconvolutional neural networkland cover |
| spellingShingle | Quang-Thanh Bui Tien-Yin Chou Thanh-Van Hoang Yao-Min Fang Ching-Yun Mu Pi-Hui Huang Vu-Dong Pham Quoc-Huy Nguyen Do Thi Ngoc Anh Van-Manh Pham Michael E. Meadows Gradient Boosting Machine and Object-Based CNN for Land Cover Classification Remote Sensing object-based image analysis gradient boosting convolutional neural network land cover |
| title | Gradient Boosting Machine and Object-Based CNN for Land Cover Classification |
| title_full | Gradient Boosting Machine and Object-Based CNN for Land Cover Classification |
| title_fullStr | Gradient Boosting Machine and Object-Based CNN for Land Cover Classification |
| title_full_unstemmed | Gradient Boosting Machine and Object-Based CNN for Land Cover Classification |
| title_short | Gradient Boosting Machine and Object-Based CNN for Land Cover Classification |
| title_sort | gradient boosting machine and object based cnn for land cover classification |
| topic | object-based image analysis gradient boosting convolutional neural network land cover |
| url | https://www.mdpi.com/2072-4292/13/14/2709 |
| work_keys_str_mv | AT quangthanhbui gradientboostingmachineandobjectbasedcnnforlandcoverclassification AT tienyinchou gradientboostingmachineandobjectbasedcnnforlandcoverclassification AT thanhvanhoang gradientboostingmachineandobjectbasedcnnforlandcoverclassification AT yaominfang gradientboostingmachineandobjectbasedcnnforlandcoverclassification AT chingyunmu gradientboostingmachineandobjectbasedcnnforlandcoverclassification AT pihuihuang gradientboostingmachineandobjectbasedcnnforlandcoverclassification AT vudongpham gradientboostingmachineandobjectbasedcnnforlandcoverclassification AT quochuynguyen gradientboostingmachineandobjectbasedcnnforlandcoverclassification AT dothingocanh gradientboostingmachineandobjectbasedcnnforlandcoverclassification AT vanmanhpham gradientboostingmachineandobjectbasedcnnforlandcoverclassification AT michaelemeadows gradientboostingmachineandobjectbasedcnnforlandcoverclassification |