There is a growing need to increase global crop yields, whilst minimising use of resources such as land, fertilisers and water. Agricultural researchers use ground-based observations to identify, select and develop crops with favourable genotypes and phenotypes; however, the ability to collect rapid, high quality and high volume phenotypic data in open fields is restricting this. This study develops and assesses a method for deriving crop height and growth rate rapidly from multi-temporal, very high spatial resolution (1 cm/pixel), 3D digital surface models of crop field trials produced via Structure from Motion (SfM) photogrammetry using aerial imagery collected through repeated campaigns flying an Unmanned Aerial Vehicle (UAV) with a mounted Red Green Blue (RGB) camera. We compare UAV SfM modelled crop heights to those derived from terrestrial laser scanner (TLS) and to the standard field measurement of crop height conducted using a 2 m rule. The most accurate UAV-derived surface model and the TLS both achieve a Root Mean Squared Error (RMSE) of 0.03 m compared to the existing manual 2 m rule method. The optimised UAV method was then applied to the growing season of a winter wheat field phenotyping experiment containing 25 different varieties grown in 27 m² plots and subject to four different nitrogen fertiliser treatments. Accuracy assessments at different stages of crop growth produced consistently low RMSE values (0.07, 0.02 and 0.03 m for May, June and July, respectively), enabling crop growth rate to be derived from differencing of the multi-temporal surface models. We find growth rates range from −13 mm/day to 17 mm/day. Our results clearly display the impact of variable nitrogen fertiliser rates on crop growth. Digital surface models produced provide a novel spatial mapping of crop height variation both at the field scale and also within individual plots. This study proves UAV based SfM has the potential to become a new standard for high-throughput phenotyping of in-field crop heights.

Background: Wheat spike architecture is a key determinant of multiple grain yield components and detailed examination of spike morphometric traits is beneficial to explain wheat grain yield and the effects of differing agronomy and genetics. However, quantification of spike morphometric traits has been very limited because it relies on time-consuming manual measurements. Results: In this study, using X-ray Computed Tomography imaging, we proposed a method to efficiently detect the 3D architecture of wheat spikes and component spikelets by clustering grains based on their Euclidean distance and relative positions. Morphometric characteristics of wheat spikelets and grains, e.g., number, size and spatial distribution along the spike can be determined. Two commercial wheat cultivars, one old, Maris Widgeon, and one modern, Siskin, were studied as examples. The average grain volume of Maris Widgeon and Siskin did not differ, but Siskin had more grains per spike and therefore greater total grain volume per spike. The spike length and spikelet number were not statistically different between the two cultivars. However, Siskin had a higher spikelet density (number of spikelets per unit spike length), with more grains and greater grain volume per spikelet than Maris Widgeon. Spatial distribution analysis revealed the number of grains, the average grain volume and the total grain volume of individual spikelets varied along the spike. Siskin had more grains and greater grain volumes per spikelet from spikelet 6, but not spikelet 1–5, compared with Maris Widgeon. The distribution of average grain volume along the spike was similar for the two wheat cultivars. Conclusion: The proposed method can efficiently extract spike, spikelet and grain morphometric traits of different wheat cultivars, which can contribute to a more detailed understanding of the sink of wheat grain yield.

Accurate detection of drought stress in plants is essential for water use efficiency and agricultural output. Hyperspectral imaging (HSI) provides a non-invasive method in plant phenotyping, allowing the long-term monitoring of plant health due to sensitivity to subtle changes in leaf constituents. The broad spectral range of HSI enables the development of different vegetation indices (VIs) to analyze plant trait responses to multiple stresses, such as the combination of nutrient and drought stresses. However, known VIs may underperform when subjected to multiple stresses. This study presents new VIs in tandem with machine learning models to identify drought stress in wheat plants under varying nitrogen (N) levels. A pot wheat experiment was set up in the glasshouse with four treatments: well-watered high-N (WWHN), well-watered low-N (WWLN), drought-stress high-N (DSHN) and drought-stress low-N (DSLN). In addition to ensuring that plants were watered according to the experiment design, photosynthetic rate (P_n) and stomatal conductance (g_s) (which are used to assess plant drought stress) were taken regularly, serving as the ground truth data for this study. The proposed VIs, together with known VIs, were used to train three classification models: support vector machines (SVM), random forest (RF), and deep neural networks (DNN) to classify plants based on their drought status. The proposed VIs achieved more than 0.94 accuracy across all models, and their performance further increased when combined with known VIs. The combined VIs were used to train three regression models to predict the stomatal conductance and photosynthetic rates of plants. The random forest regression model performed best, suggesting that it could be used as a stand-alone tool to forecast g_s and P_n and track drought stress in wheat. This study shows that combining hyperspectral data with machine learning can effectively monitor and predict drought stress in crops, especially in varying nitrogen conditions.

Vegetation indices, such as the Normalised Difference Vegetation Index (NDVI), are common metrics used for measuring traits of interest in crop phenotyping. However, traditional measurements of these indices are often influenced by multiple confounding factors such as canopy cover and reflectance of underlying soil, visible in canopy gaps. Digital cameras mounted to Unmanned Aerial Vehicles offer the spatial resolution to investigate these confounding factors, however incomplete methods for radiometric calibration into reflectance units limits how the data can be applied to phenotyping. In this study, we assess the applicability of very high spatial resolution (1 cm) UAV-based imagery taken with commercial off the shelf (COTS) digital cameras for both deriving calibrated reflectance imagery, and isolating vegetation canopy reflectance from that of the underlying soil. We present new methods for successfully normalising COTS camera imagery for exposure and solar irradiance effects, generating multispectral (RGB-NIR) orthomosaics of our target field-based wheat crop trial. Validation against measurements from a ground spectrometer showed good results for reflectance (R² ≥ 0.6) and NDVI (R² ≥ 0.88). Application of imagery collected through the growing season and masked using the Excess Green Red index was used to assess the impact of canopy cover on NDVI measurements. Results showed the impact of canopy cover artificially reducing plot NDVI values in the early season, where canopy development is low.

Image segmentation is a fundamental but critical step for achieving automated high- throughput phenotyping. While conventional segmentation methods perform well in homogenous environments, the performance decreases when used in more complex environments. This study aimed to develop a fast and robust neural-network-based segmentation tool to phenotype plants in both field and glasshouse environments in a high-throughput manner. Digital images of cowpea (from glasshouse) and wheat (from field) with different nutrient supplies across their full growth cycle were acquired. Image patches from 20 randomly selected images from the acquired dataset were transformed from their original RGB format to multiple color spaces. The pixels in the patches were annotated as foreground and background with a pixel having a feature vector of 24 color properties. A feature selection technique was applied to choose the sensitive features, which were used to train a multilayer perceptron network (MLP) and two other traditional machine learning models: support vector machines (SVMs) and random forest (RF). The performance of these models, together with two standard color-index segmentation techniques (excess green (ExG) and excess green–red (ExGR)), was compared. The proposed method outperformed the other methods in producing quality segmented images with over 98%-pixel classification accuracy. Regression models developed from the different segmentation methods to predict Soil Plant Analysis Development (SPAD) values of cowpea and wheat showed that images from the proposed MLP method produced models with high predictive power and accuracy comparably. This method will be an essential tool for the development of a data analysis pipeline for high-throughput plant phenotyping. The proposed technique is capable of learning from different environmental conditions, with a high level of robustness.

Due to the adverse effect of prolonged drought stress on plants, accurate detection is essential for water use efficiency and maintaining productivity. Hyperspectral imaging is frequently used for non-invasive plant phenotyping, allowing for the long-term monitoring of crop health due to its sensitivity to subtle changes in leaf constituents. The broad spectrum of hyperspectral data enables the development of multiple vegetation indices (Vis) derived from the different spectral regions to estimate plant biophysical and biochemical traits. However, the known VIs often do not generalize well and perform poorly for multiple plant stresses. This study proposes new VIs combined with machine learning models to identify drought stress in wheat species under different nitrogen (N) levels. A wheat experiment was set up in the glasshouse with four treatments: well-watered high-N (WWHN), well-watered low-N (WWLN), drought-stress high N and drought-stress low-N. In addition to ensuring that plants were watered according to the experiment design, photosynthetic rate (Pn) and stomatal conductance (gs) were taken regularly, serving as the ground truth data for this study. Sensitive spectral features were selected using a custom-designed ensemble modelling technique. New drought VIs are proposed using different combinations of the selected features. Three classification models (support vector machines, random forest and deep neural network) were developed and trained using four sets of data: known VIs, proposed VIs, combined VIs (from the known and proposed VIs) and PCA-transformed features (over the whole spectral data). From the results, the proposed VIs outperformed the known VIs, yielding > 0.94 accuracies for all three models, and the performance improved when they were trained with the combined VIs. The combined VIs were used to train three regression models to predict the stomatal conductances and photosynthetic rates of plants. The random forest regression model performed best, suggesting that it could be used as a stand-alone tool to forecast gs and Pn and track drought stress in wheat.