Combining Deep Learning and Prior Knowledge for Crop Mapping in Tropical Regions from Multitemporal SAR Image Sequences
Accurate crop type identification and crop area estimation from remote sensing data in tropical regions are still considered challenging tasks. The more favorable weather conditions, in comparison to the characteristic conditions of temperate regions, permit higher flexibility in land use, planning,...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2019-08-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/11/17/2029 |
| _version_ | 1819110873929613312 |
|---|---|
| author | Laura Elena Cué La Rosa Raul Queiroz Feitosa Patrick Nigri Happ Ieda Del’Arco Sanches Gilson Alexandre Ostwald Pedro da Costa |
| author_facet | Laura Elena Cué La Rosa Raul Queiroz Feitosa Patrick Nigri Happ Ieda Del’Arco Sanches Gilson Alexandre Ostwald Pedro da Costa |
| author_sort | Laura Elena Cué La Rosa |
| collection | DOAJ |
| description | Accurate crop type identification and crop area estimation from remote sensing data in tropical regions are still considered challenging tasks. The more favorable weather conditions, in comparison to the characteristic conditions of temperate regions, permit higher flexibility in land use, planning, and management, which implies complex crop dynamics. Moreover, the frequent cloud cover prevents the use of optical data during large periods of the year, making SAR data an attractive alternative for crop mapping in tropical regions. This paper evaluates the effectiveness of Deep Learning (DL) techniques for crop recognition from multi-date SAR images from tropical regions. Three DL strategies are investigated: autoencoders, convolutional neural networks, and fully-convolutional networks. The paper further proposes a post-classification technique to enforce prior knowledge about crop dynamics in the target area. Experiments conducted on a Sentinel-1 multitemporal sequence of a tropical region in Brazil reveal the pros and cons of the tested methods. In our experiments, the proposed crop dynamics model was able to correct up to 16.5% of classification errors and managed to improve the performance up to 3.2% and 8.7% in terms of overall accuracy and average F1-score, respectively. |
| first_indexed | 2024-12-22T03:48:39Z |
| format | Article |
| id | doaj.art-35eed255c0bc4dcc98a283b9fd09e0f3 |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-12-22T03:48:39Z |
| publishDate | 2019-08-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-35eed255c0bc4dcc98a283b9fd09e0f32022-12-21T18:40:04ZengMDPI AGRemote Sensing2072-42922019-08-011117202910.3390/rs11172029rs11172029Combining Deep Learning and Prior Knowledge for Crop Mapping in Tropical Regions from Multitemporal SAR Image SequencesLaura Elena Cué La Rosa0Raul Queiroz Feitosa1Patrick Nigri Happ2Ieda Del’Arco Sanches3Gilson Alexandre Ostwald Pedro da Costa4Department of Electrical Engineering, Pontifical Catholic University of Rio de Janeiro, Rio de Janeiro 22451-900, Rio de Janeiro, BrazilDepartment of Electrical Engineering, Pontifical Catholic University of Rio de Janeiro, Rio de Janeiro 22451-900, Rio de Janeiro, BrazilDepartment of Electrical Engineering, Pontifical Catholic University of Rio de Janeiro, Rio de Janeiro 22451-900, Rio de Janeiro, BrazilNational Institute for Space Research (INPE), São Jose dos Campos 12227-010, São Paulo, BrazilDepartment of Informatics and Computer Science, Rio de Janeiro State University, Rio de Janeiro 20550-900, Rio de Janeiro, BrazilAccurate crop type identification and crop area estimation from remote sensing data in tropical regions are still considered challenging tasks. The more favorable weather conditions, in comparison to the characteristic conditions of temperate regions, permit higher flexibility in land use, planning, and management, which implies complex crop dynamics. Moreover, the frequent cloud cover prevents the use of optical data during large periods of the year, making SAR data an attractive alternative for crop mapping in tropical regions. This paper evaluates the effectiveness of Deep Learning (DL) techniques for crop recognition from multi-date SAR images from tropical regions. Three DL strategies are investigated: autoencoders, convolutional neural networks, and fully-convolutional networks. The paper further proposes a post-classification technique to enforce prior knowledge about crop dynamics in the target area. Experiments conducted on a Sentinel-1 multitemporal sequence of a tropical region in Brazil reveal the pros and cons of the tested methods. In our experiments, the proposed crop dynamics model was able to correct up to 16.5% of classification errors and managed to improve the performance up to 3.2% and 8.7% in terms of overall accuracy and average F1-score, respectively.https://www.mdpi.com/2072-4292/11/17/2029crop mappingtropical agricultureSARdeep learningSentinel-1multitemporal image analysis |
| spellingShingle | Laura Elena Cué La Rosa Raul Queiroz Feitosa Patrick Nigri Happ Ieda Del’Arco Sanches Gilson Alexandre Ostwald Pedro da Costa Combining Deep Learning and Prior Knowledge for Crop Mapping in Tropical Regions from Multitemporal SAR Image Sequences Remote Sensing crop mapping tropical agriculture SAR deep learning Sentinel-1 multitemporal image analysis |
| title | Combining Deep Learning and Prior Knowledge for Crop Mapping in Tropical Regions from Multitemporal SAR Image Sequences |
| title_full | Combining Deep Learning and Prior Knowledge for Crop Mapping in Tropical Regions from Multitemporal SAR Image Sequences |
| title_fullStr | Combining Deep Learning and Prior Knowledge for Crop Mapping in Tropical Regions from Multitemporal SAR Image Sequences |
| title_full_unstemmed | Combining Deep Learning and Prior Knowledge for Crop Mapping in Tropical Regions from Multitemporal SAR Image Sequences |
| title_short | Combining Deep Learning and Prior Knowledge for Crop Mapping in Tropical Regions from Multitemporal SAR Image Sequences |
| title_sort | combining deep learning and prior knowledge for crop mapping in tropical regions from multitemporal sar image sequences |
| topic | crop mapping tropical agriculture SAR deep learning Sentinel-1 multitemporal image analysis |
| url | https://www.mdpi.com/2072-4292/11/17/2029 |
| work_keys_str_mv | AT lauraelenacuelarosa combiningdeeplearningandpriorknowledgeforcropmappingintropicalregionsfrommultitemporalsarimagesequences AT raulqueirozfeitosa combiningdeeplearningandpriorknowledgeforcropmappingintropicalregionsfrommultitemporalsarimagesequences AT patricknigrihapp combiningdeeplearningandpriorknowledgeforcropmappingintropicalregionsfrommultitemporalsarimagesequences AT iedadelarcosanches combiningdeeplearningandpriorknowledgeforcropmappingintropicalregionsfrommultitemporalsarimagesequences AT gilsonalexandreostwaldpedrodacosta combiningdeeplearningandpriorknowledgeforcropmappingintropicalregionsfrommultitemporalsarimagesequences |