Paddy Rice Maps South Korea (2018)

Description

This dataset includes three paddy rice maps in South Korea for the year 2018. Specifically, the three maps are for the regions of i) Seosan and Dangjin, ii) Haenam and iii) Cheorwon. The paddy rice maps are a product Random Forest predictions and DO NOT represent ground truth information. The predictions were made based on a semi-supervised approach that first generates pseudo-labels using a dynamic k-means algorithm and then trains the Random Forest Classifier. The classifier has been trained on pseudo-labels that stem from the Seosan and Dangjin site (i) and fine-tuned for generalization using labeled samples from sites (ii) and (iii). For more information please read the publication "A scalable machine learning pipeline for paddy rice classification using multi-temporal Sentinel data" in MDPI Remote Sensing (pending).

Citation

Please cite this paper: Sitokonstantinou V, Koukos A, Drivas T, Kontoes C, Papoutsis I, Karathanassi V. A Scalable Machine Learning Pipeline for Paddy Rice Classification Using Multi-Temporal Sentinel Data. Remote Sensing. 2021; 13(9):1769. https://doi.org/10.3390/rs13091769

• Files The map can be found here or

  i) Seosan-Dangjin Map

  ii) Haenam Map

  iii) Cheorwon Map