The Application of an Unmanned Aerial System and Machine Learning Techniques for Red Clover-Grass Mixture Yield Estimation under Variety Performance Trials

Abstract

A significant trend has developed with the recent growing interest in the estimation of aboveground biomass of vegetation in legume-supported systems in perennial or semi-natural grasslands to meet the demands of sustainable and precise agriculture. Unmanned aerial systems (UAS) are a powerful tool when it comes to supporting farm-scale phenotyping trials. In this study, we explored the variation of the red clover-grass mixture dry matter (DM) yields between temporal periods (one- and two-year cultivated), farming operations [soil tillage methods (STM), cultivation methods (CM), manure application (MA)] using three machine learning (ML) techniques [random forest regression (RFR), support vector regression (SVR), and artificial neural network (ANN)] and six multispectral vegetation indices (VIs) to predict DM yields. The ML evaluation results showed the best performance for ANN in the 11-day before harvest category (R² = 0.90, NRMSE = 0.12), followed by RFR (R² = 0.90 NRMSE = 0.15), and SVR (R² = 0.86, NRMSE = 0.16), which was furthermore supported by the leave-one-out cross-validation pre-analysis. In terms of VI performance, green normalized difference vegetation index (GNDVI), green difference vegetation index (GDVI), as well as modified simple ratio (MSR) performed better as predictors in ANN and RFR. However, the prediction ability of models was being influenced by farming operations. The stratified sampling, based on STM, had a better model performance than CM and MA. It is proposed that drone data collection was suggested to be optimum in this study, closer to the harvest date, but not later than the ageing stage.

An Overview of Machine Learning Applications on Plant Phenotyping, with a Focus on Sunflower

“…The Deep Neural Network (DNN), a more robust and sophisticated model, was used to analyse RGB, multispectral, and thermal images to predict soybean yield [126]. Hyperspectral imaging combined with ML was also applied to classify and predict plants traits, such as salt stress [130], crop yield [131][132][133][134][135][136], and biomass quantity [137][138][139][140]. However, powerful data mining techniques are required.…”

Toward Automated Machine Learning-Based Hyperspectral Image Analysis in Crop Yield and Biomass Estimation

“…The assessment was carried out for the prediction of AutoML models (Figures 6 and 7). Performance evaluation approaches proposed by [19,93] were utilized to evaluate each model. The coefficient of determination (R 2 ) (Equation (1) and normalized root means square error (NRMSE) (Equation ( 2)) were used to evaluate the models' accuracy.…”

“…The use of RS technologies provides timely, non-destructive, spatial estimates for measuring and tracking specific vegetation attributes [7], as well as continuing to improve crop yield production and quality, thereby assisting in future food security and reducing the negative impacts of agricultural practices [5,8,9]. Moreover, agriculture management practices based on the concept of sustainable cropping ideas (such as reduced tillage intensity [10][11][12][13][14][15], fertilizer input [16], and organic farming [17,18]) combined with mixed cropping systems, particularly legume-based, can effectively diminish greenhouse gas emissions by reducing the use of inorganic nitrogen fertilizers and replacing them with symbiotically fixed nitrogen [19], as well as carbon loss [5,20,21] and soil erosion [22] in cultivated soil. Furthermore, they can contribute to productivity and economic appeal to Northern European farmers, which is crucial for ensuring that these ecologically friendly systems can compete in terms of profitability with more traditional or artificially generated systems [23].…”

A Systematic Review of UAV Applications for Mapping Neglected and Underutilised Crop Species’ Spatial Distribution and Health

“…Inherent water scarcity, which is exacerbated by factors such as climate change, population expansion, and changes in land use [1][2][3], has intensified the pressure on the agricultural sector, particularly in ensuring long-term food supply for the expanding populations [4,5]. Most of the agricultural production in developing regions is derived from rainfed farms, which occupy 97% of croplands [6]. However, 80% of these croplands are smallholder farms that contribute most of the food production in developing regions [1,2].…”

“…RS and ML techniques have recently greatly aided high-throughput phenotyping technologies [12]. For example Li, Burnside [6] combined three ML techniques, which included random forest (RFR), support vector regression (SVR), and artificial neural network (ANN), in combination with optimal VI's to predict the red-clover dry matter yields in various phenological growth periods.…”