Combining Deep Learning and Prior Knowledge for Crop Mapping in Tropical Regions from Multitemporal SAR Image Sequences

• Published in: Remote sensing (Basel, Switzerland) Vol. 11; no. 17; p. 2029
• Main Authors: Cué La Rosa, Laura Elena, Queiroz Feitosa, Raul, Nigri Happ, Patrick, Del’Arco Sanches, Ieda, Ostwald Pedro da Costa, Gilson Alexandre
• Format: Journal Article
• Language: English
• Published: Basel MDPI AG 2019
• Subjects: Agricultural management; Agricultural production; Agriculture; Artificial neural networks; Classification; Cloud cover; crop mapping; Crop rotation; Crops; Deep learning; Earth science; Food security; Freshwater resources; Irrigation; Land use; Land use management; Machine learning; Mapping; multitemporal image analysis; Neural networks; Object recognition; Performance enhancement; Remote sensing; SAR; Semantics; Sentinel-1; Synthetic aperture radar; tropical agriculture; Tropical environment; Tropical environments; Vegetation mapping; Weather; Brazil; Rio de Janeiro Brazil; United States > US; Ukraine
• ISSN: 2072-4292; 2072-4292
• DOI: 10.3390/rs11172029

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.