A Novel Rice Mapping Method Based on Multi-Temporal Polarization Decomposition Components for Single-Season Rice

Currently, Synthetic Aperture Radar (SAR) time-series data are more widely used in rice mapping for their ability to work all-day and all-weather. However, most researches only use backscattering coefficients while neglecting polarimetric information in SAR data. Polarimetric information can characterize the scattering mechanism of ground objects and thus greatly simplify the mapping task. To address this issue, this study proposes a feature combination based on the m/X decomposition method and the classic UNet model to extract rice planting areas with Sentinel-1 dual-polarized data. The training accuracy on validation set can reach 98.7%, and the mIoU and mPA of rice planting areas achieve 91.74% and 95.66%, respectively. These experimental results indicate that the proposed method is simple and effective, and can be generalized for large-scale rice mapping.