Rice-Yield Prediction with Multi-Temporal Sentinel-2 Data and 3D CNN: A Case Study in Nepal

Abstract

Crop yield estimation is a major issue of crop monitoring which remains particularly challenging in developing countries due to the problem of timely and adequate data availability. Whereas traditional agricultural systems mainly rely on scarce ground-survey data, freely available multi-temporal and multi-spectral remote sensing images are excellent tools to support these vulnerable systems by accurately monitoring and estimating crop yields before harvest. In this context, we introduce the use of Sentinel-2 (S2) imagery, with a medium spatial, spectral and temporal resolutions, to estimate rice crop yields in Nepal as a case study. Firstly, we build a new large-scale rice crop database (RicePAL) composed by multi-temporal S2 and climate/soil data from the Terai districts of Nepal. Secondly, we propose a novel 3D Convolutional Neural Network (CNN) adapted to these intrinsic data constraints for the accurate rice crop yield estimation. Thirdly, we study the effect of considering different temporal, climate and soil data configurations in terms of the performance achieved by the proposed approach and several state-of-the-art regression and CNN-based yield estimation methods. The extensive experiments conducted in this work demonstrate the suitability of the proposed CNN-based framework for rice crop yield estimation in the developing country of Nepal using S2 data.

Development of Crop-Weather Models Using Gaussian Process Regression for the Prediction of Paddy Yield in Sri Lanka

“…If one of the Kernel functions recommended in previous research studies (Exponential/ RBF /Squared Exponential) were applied, the accuracy of the proposed crop-weather models would be reduced. More importantly, the Matern 5/2 Kernel, which was the most suitable function in developing models for 6 regions in this research, was not recommended in previous related research studies [16][17][18][19]. The use of Exponential Kernel, which was found to be the best in developing models for the other 3 regions, was also not recommended by other researchers though it was used in previous research conducted by the authors [11].…”

“…Though they had used only the Radial Basis Function (RBF) and the Rational Quadratic Kernel, GPR and DNNR were reported to have performed well with R2 > 0.90. According to research on the paddy yield estimation in Nepal, GPR outperformed SVMR, Linear Regression, and Ridge Regression [18]. However, this study was limited to a single Kernel of Squared Exponential.…”

Quantifying the Cooling Effect and Scale of Large Inner-City Lakes Based on Landscape Patterns: A Case Study of Hangzhou and Nanjing

“…The 20 m bands were up-sampled to 10 m resolution using nearest-neighbor interpolation [39]. Bands 1, 9, and 10 at 60 m spatial resolution, dedicated to atmospheric correction and cirrus detection, were discarded [40,41].…”

Rice Mapping in a Subtropical Hilly Region Based on Sentinel-1 Time Series Feature Analysis and the Dual Branch BiLSTM Model

“…In recent years, deep learning models, such as convolutional neural networks (CNNs) and recurrent neural networks (RNNs), have provided strong technical support for crop mapping [33,34]. Kussul et al used 19 scenes of Sentinel-1A and Landsat-8 data and a shallow CNN for crop classification in Ukraine, which showed that the CNN architecture outperformed the multilayer perceptron architecture, resulting in accuracies above 85% for major crops [35].…”