Mapping Paddy Rice Using a Convolutional Neural Network (CNN) with Landsat 8 Datasets in the Dongting Lake Area, China

TLDR

Rice is a major staple food worldwide, especially in China, and accurate monitoring of rice‑producing land is essential for assessing food supplies and productivity, with deep‑learning convolutional neural networks recently proving effective in remote‑sensing analysis. This study develops a CNN‑based paddy‑rice mapping method that integrates multitemporal Landsat 8 imagery, phenological data, and land‑surface temperature. The method first fuses MODIS and Landsat data with STARFM to generate multitemporal Landsat‑like images, then extracts phenological variables from NDVI time series using a threshold approach, derives LST via a generalized single‑channel algorithm, and finally applies a patch‑based deep‑learning CNN to the combined spectral, phenology, and LST data to extract paddy‑rice information. The proposed approach achieved an overall accuracy of 97.06 % and a Kappa of 0.91, outperforming SVM and random‑forest baselines by 6.43 %/0.07 and 7.68 %/0.09 respectively, and its Landsat‑derived rice area correlates strongly (R² = 0.9945) with government statistics, indicating strong potential for large‑scale mapping.

Abstract

Rice is one of the world’s major staple foods, especially in China. Highly accurate monitoring on rice-producing land is, therefore, crucial for assessing food supplies and productivity. Recently, the deep-learning convolutional neural network (CNN) has achieved considerable success in remote-sensing data analysis. A CNN-based paddy-rice mapping method using the multitemporal Landsat 8, phenology data, and land-surface temperature (LST) was developed during this study. First, the spatial–temporal adaptive reflectance fusion model (STARFM) was used to blend the moderate-resolution imaging spectroradiometer (MODIS) and Landsat data for obtaining multitemporal Landsat-like data. Subsequently, the threshold method is applied to derive the phenological variables from the Landsat-like (Normalized difference vegetation index) NDVI time series. Then, a generalized single-channel algorithm was employed to derive LST from the Landsat 8. Finally, multitemporal Landsat 8 spectral images, combined with phenology and LST data, were employed to extract paddy-rice information using a patch-based deep-learning CNN algorithm. The results show that the proposed method achieved an overall accuracy of 97.06% and a Kappa coefficient of 0.91, which are 6.43% and 0.07 higher than that of the support vector machine method, and 7.68% and 0.09 higher than that of the random forest method, respectively. Moreover, the Landsat-derived rice area is strongly correlated (R2 = 0.9945) with government statistical data, demonstrating that the proposed method has potential in large-scale paddy-rice mapping using moderate spatial resolution images.