Farm: A Fully Automated Rice Mapping Framework Combining Sentinel-1 Sar and Sentinel-2 Multi-Temporal Imagery

Abstract

Accurately and timely acquiring the spatial planting structure of paddy rice over large regions is a prerequisite for ensuring national food security and exploring the impacts of climate change. To date, such studies have been limited by the difficulty of obtaining optical cloud-free data and by the spackle noise of radar data; many challenges arise when identifying crops using only a single data source. Additionally, obtaining a sufficient quantity of representative training samples that accurately reflect diverse phenological patterns is a challenging task for conducting large-scale monitoring and classification of rice cultivation. To overcome these challenges, this study proposed a fully automated rice-mapping frame (FARM) combining the advantages of time-series synthetic aperture radar (SAR) and optical satellite imagery to enable large-area rice mapping without any manually collected samples. First, an object-based fully automatic training sample generation strategy was proposed. Through the use of special rice-flooding features, based on the time-series SAR satellite images, the phenology constraint rule is constructed to extract the rice sample object. Second, the rice sample objects extracted based on phenological rules were used as training samples for paddy rice extraction by integrating multiple random forest (RF) classifiers, referred to as the multi-RF method, in which each RF is built individually based on images acquired in each phenological phase of the growing season. Furthermore, the study explored the availability of the method in early-season rice identification by transferring the training samples acquired by the FARM to a new year. The proposed FARM approach was then validated under different cropping conditions at three study sites in China. The results show that the FARM framework proved to be more effective than other methods at all three study sites, achieving average overall accuracies (OA) ranging from 92.80% - 97.00%. In addition, when transferring the training samples from 2021 to other years (2020/2022), the OAs of site A, site B and site C were high during the heading period, with accuracies of 97.57%, 84.28% and 89.27%, respectively. These results demonstrate that, first, the FARM framework exhibits high efficiency and accuracy in different study areas without the need for extensive fieldwork to collect training samples. Second, the method has good performance in the early-season rice mapping of the new year and can be used to perform timely and accurate rice identification and monitoring tasks. The method shows great potential in obtaining large-area automatic rice mapping results in a timely and accurate manner.

Keywords: rice identification, training sample generation, Image segmentation, phenology, time-series satellite imagery