Search

High-throughput field phenotyping using close remote sensing platforms and sensors for non-destructive assessment of plant traits can support the objective evaluation of yield predictions of large breeding trials. The main objective of this study was to examine the potential of unmanned aerial vehicle (UAV)-based structural and spectral features and their combination in herbage yield predictions across diploid and tetraploid varieties and breeding populations of perennial ryegrass (Lolium perenne L.). Canopy structural (i.e., canopy height) and spectral (i.e., vegetation indices) information were derived from data gathered with two sensors: a consumer-grade RGB and a 10-band multispectral (MS) camera system, which were compared in the analysis. A total of 468 field plots comprising 115 diploid and 112 tetraploid varieties and populations were considered in this study. A modelling framework established to predict dry matter yield (DMY), was used to test three machine learning algorithms, including Partial Least Squares Regression (PLSR), Random Forest (RF), and Support Vector Machines (SVM). The results of the nested cross-validation revealed: (a) the fusion of structural and spectral features achieved better DMY estimates as compared to models fitted with structural or spectral data only, irrespective of the sensor, ploidy level or machine learning algorithm applied; (b) models built with MS-based predictor variables, despite their lower spatial resolution, slightly outperformed the RGB-based models, as lower mean relative root mean square error (rRMSE) values were delivered; and (c) on average, the RF technique reported the best model performances among tested algorithms, regardless of the dataset used. The approach introduced in this study can provide accurate yield estimates (up to an RMSE = 308 kg ha−1) and useful information for breeders and practical farm-scale applications. [ABSTRACT FROM AUTHOR]

The development of approaches to determine the spatial variability of nitrogen (N) into coffee leaves is essential to increase productivity and reduce production costs and environmental impacts associated with excessive N applications. Thus, this study aimed to assess the potential of the Random Forest (RF) machine learning method applied to vegetation indices (VI) obtained from Remotely Piloted Aircraft (RPA) images to measure the N content in coffee plants. A total of 10 VI were obtained from multispectral images by a camera attached to a rotary-wing RPA. The RGB orthomosaic was used to determine sampling points at the crop area, which were ranked by N levels in the plants as deficient, critical, or sufficient. The chemical analysis of N content in the coffee leaves, as well as the VI values in sample points, were used as input parameters for the image training and its classification by the RF. The suggested model has shown global accuracy and a kappa coefficient of up to 0.91 and 0.86, respectively. The best results were achieved using the Green Normalized Difference Vegetation (GNDVI) and Green Optimized Soil Adjusted Vegetation Index (GOSAVI). In addition, the model enabled the evaluation of the spatial distribution of N in the coffee trees, as well as quantification of N deficiency in the crop for the whole area. The GNDVI and GOSAVI allowed the verification that 22% of the entire crop area had plants with N deficiency symptoms, which would result in a reduction of 78% in the amount of N applied by the producer. [ABSTRACT FROM AUTHOR]