Validation of UAV-based alfalfa biomass predictability using photogrammetry with fully automatic plot segmentation

Phenotyping Alfalfa (Medicago sativa L.) Root Structure Architecture via Integrating Confident Machine Learning with ResNet-18

Background: Root system architecture (RSA) is of growing interest in implementing plant improvements with belowground root traits. Modern computing technology applied to images offers new pathways forward to plant trait improvements and selection through RSA analysis (using images to discern/classify root types and traits). However, a major stumbling block to image-based RSA phenotyping is image label noise, which reduces the accuracies of models that take images as direct inputs. To address the label noise problem, this study utilized an artificial intelligence model capable of classifying the RSA of alfalfa (Medicago sativa L.) directly from images and coupled it with downstream label improvement methods. Images were compared with different model outputs with manual root classifications, and confident machine learning (CL) and reactive machine learning (RL) methods were tested to minimize the effects of subjective labeling to improve labeling and prediction accuracies. Results: The CL algorithm modestly improved the Random Forest model's overall prediction accuracy of the Minnesota dataset (1%) while larger gains in accuracy were observed with the ResNet-18 model results. The ResNet-18 cross-population prediction accuracy was improved (~8% to 13%) with CL compared to the original/preprocessed datasets. Training and testing data combinations with the highest accuracies (86%) resulted from the CL- and/or RL-corrected datasets for predicting taproot RSAs. Similarly, the highest accuracies achieved for the intermediate RSA class resulted from corrected data combinations. The highest overall accuracy (~75%) using the ResNet-18 model involved CL on a pooled dataset containing images from both sample locations. Conclusions: ResNet-18 DNN prediction accuracies of alfalfa RSA image labels are increased when CL and RL are employed. By increasing the dataset to reduce overfitting while concurrently finding and correcting image label errors, it is demonstrated here that accuracy increases by as much as ~11% to 13% can be achieved with semi-automated, computer-assisted preprocessing and data cleaning (CL/RL).

The State of the Art in Root System Architecture Image Analysis Using Artificial Intelligence: A Review

Roots are essential for acquiring water and nutrients to sustain and support plant growth and anchorage. However, they have been studied less than the aboveground traits in phenotyping and plant breeding until recent decades. In modern times, root properties such as morphology and root system architecture (RSA) have been recognized as increasingly important traits for creating more and higher quality food in the "Second Green Revolution". To address the paucity in RSA and other root research, new technologies are being investigated to fill the increasing demand to improve plants via root traits and overcome currently stagnated genetic progress in stable yields. Artificial intelligence (AI) is now a cutting-edge technology proving to be highly successful in many applications, such as crop science and genetic research to improve crop traits. A burgeoning field in crop science is the application of AI to high-resolution imagery in analyses that aim to answer questions related to crops and to better and more speedily breed desired plant traits such as RSA into new cultivars. This review is a synopsis concerning the origins, applications, challenges, and future directions of RSA research regarding image analyses using AI.

Utilizing Spectral, Structural and Textural Features for Estimating Oat Above-Ground Biomass Using UAV-Based Multispectral Data and Machine Learning

Accurate and timely monitoring of biomass in breeding nurseries is essential for evaluating plant performance and selecting superior genotypes. Traditional methods for phenotyping above-ground biomass in field conditions requires significant time, cost, and labor. Unmanned Aerial Vehicles (UAVs) offer a rapid and non-destructive approach for phenotyping multiple field plots at a low cost. While Vegetation Indices (VIs) extracted from remote sensing imagery have been widely employed for biomass estimation, they mainly capture spectral information and disregard the 3D canopy structure and spatial pixel relationships. Addressing these limitations, this study, conducted in 2020 and 2021, aimed to explore the potential of integrating UAV multispectral imagery-derived canopy spectral, structural, and textural features with machine learning algorithms for accurate oat biomass estimation. Six oat genotypes planted at two seeding rates were evaluated in two South Dakota locations at multiple growth stages. Plot-level canopy spectral, structural, and textural features were extracted from the multispectral imagery and used as input variables for three machine learning models: Partial Least Squares Regression (PLSR), Support Vector Regression (SVR), and Random Forest Regression (RFR). The results showed that (1) in addition to canopy spectral features, canopy structural and textural features are also important indicators for oat biomass estimation; (2) combining spectral, structural, and textural features significantly improved biomass estimation accuracy over using a single feature type; (3) machine learning algorithms showed good predictive ability with slightly better estimation accuracy shown by RFR (R2 = 0.926 and relative root mean square error (RMSE%) = 15.97%). This study demonstrated the benefits of UAV imagery-based multi-feature fusion using machine learning for above-ground biomass estimation in oat breeding nurseries, holding promise for enhancing the efficiency of oat breeding through UAV-based phenotyping and crop management practices.

UAV-based individual Chinese cabbage weight prediction using multi-temporal data

The use of unmanned aerial vehicles (UAVs) has facilitated crop canopy monitoring, enabling yield prediction by integrating regression models. However, the application of UAV-based data to individual-level harvest weight prediction is limited by the effectiveness of obtaining individual features. In this study, we propose a method that automatically detects and extracts multitemporal individual plant features derived from UAV-based data to predict harvest weight. We acquired data from an experimental field sown with 1196 Chinese cabbage plants, using two cameras (RGB and multi-spectral) mounted on UAVs. First, we used three RGB orthomosaic images and an object detection algorithm to detect more than 95% of the individual plants. Next, we used feature selection methods and five different multi-temporal resolutions to predict individual plant weights, achieving a coefficient of determination (R2) of 0.86 and a root mean square error (RMSE) of 436 g/plant. Furthermore, we achieved predictions with an R2 greater than 0.72 and an RMSE less than 560 g/plant up to 53 days prior to harvest. These results demonstrate the feasibility of accurately predicting individual Chinese cabbage harvest weight using UAV-based data and the efficacy of utilizing multi-temporal features to predict plant weight more than one month prior to harvest.

High-precision plant height measurement by drone with RTK-GNSS and single camera for real-time processing

Conventional crop height measurements performed using aerial drone images require 3D reconstruction results of several aerial images obtained through structure from motion. Therefore, they require extensive computation time and their measurement accuracy is not high; if the 3D reconstruction result fails, several aerial photos must be captured again. To overcome these challenges, this study proposes a high-precision measurement method that uses a drone equipped with a monocular camera and real-time kinematic global navigation satellite system (RTK-GNSS) for real-time processing. This method performs high-precision stereo matching based on long-baseline lengths (approximately 1 m) during the flight by linking the RTK-GNSS and aerial image capture points. As the baseline length of a typical stereo camera is fixed, once the camera is calibrated on the ground, it does not need to be calibrated again during the flight. However, the proposed system requires quick calibration in flight because the baseline length is not fixed. A new calibration method that is based on zero-mean normalized cross-correlation and two stages least square method, is proposed to further improve the accuracy and stereo matching speed. The proposed method was compared with two conventional methods in natural world environments. It was observed that error rates reduced by 62.2% and 69.4%, for flight altitudes between 10 and 20 m respectively. Moreover, a depth resolution of 1.6 mm and reduction of 44.4% and 63.0% in the error rates were achieved at an altitude of 4.1 m, and the execution time was 88 ms for images with a size of 5472 × 3468 pixels, which is sufficiently fast for real-time measurement.

OMICS in Fodder Crops: Applications, Challenges, and Prospects

Biomass yield and quality are the primary targets in forage crop improvement programs worldwide. Low-quality fodder reduces the quality of dairy products and affects cattle's health. In multipurpose crops, such as maize, sorghum, cowpea, alfalfa, and oat, a plethora of morphological and biochemical/nutritional quality studies have been conducted. However, the overall growth in fodder quality improvement is not on par with cereals or major food crops. The use of advanced technologies, such as multi-omics, has increased crop improvement programs manyfold. Traits such as stay-green, the number of tillers per plant, total biomass, and tolerance to biotic and/or abiotic stresses can be targeted in fodder crop improvement programs. Omic technologies, namely genomics, transcriptomics, proteomics, metabolomics, and phenomics, provide an efficient way to develop better cultivars. There is an abundance of scope for fodder quality improvement by improving the forage nutrition quality, edible quality, and digestibility. The present review includes a brief description of the established omics technologies for five major fodder crops, i.e., sorghum, cowpea, maize, oats, and alfalfa. Additionally, current improvements and future perspectives have been highlighted.

AirMeasurer: open-source software to quantify static and dynamic traits derived from multiseason aerial phenotyping to empower genetic mapping studies in rice

Low-altitude aerial imaging, an approach that can collect large-scale plant imagery, has grown in popularity recently. Amongst many phenotyping approaches, unmanned aerial vehicles (UAVs) possess unique advantages as a consequence of their mobility, flexibility and affordability. Nevertheless, how to extract biologically relevant information effectively has remained challenging. Here, we present AirMeasurer, an open-source and expandable platform that combines automated image analysis, machine learning and original algorithms to perform trait analysis using 2D/3D aerial imagery acquired by low-cost UAVs in rice (Oryza sativa) trials. We applied the platform to study hundreds of rice landraces and recombinant inbred lines at two sites, from 2019 to 2021. A range of static and dynamic traits were quantified, including crop height, canopy coverage, vegetative indices and their growth rates. After verifying the reliability of AirMeasurer-derived traits, we identified genetic variants associated with selected growth-related traits using genome-wide association study and quantitative trait loci mapping. We found that the AirMeasurer-derived traits had led to reliable loci, some matched with published work, and others helped us to explore new candidate genes. Hence, we believe that our work demonstrates valuable advances in aerial phenotyping and automated 2D/3D trait analysis, providing high-quality phenotypic information to empower genetic mapping for crop improvement.

Spatial Regression Models for Field Trials: A Comparative Study and New Ideas

Naturally occurring variability within a study region harbors valuable information on relationships between biological variables. Yet, spatial patterns within these study areas, e.g., in field trials, violate the assumption of independence of observations, setting particular challenges in terms of hypothesis testing, parameter estimation, feature selection, and model evaluation. We evaluate a number of spatial regression methods in a simulation study, including more realistic spatial effects than employed so far. Based on our results, we recommend generalized least squares (GLS) estimation for experimental as well as for observational setups and demonstrate how it can be incorporated into popular regression models for high-dimensional data such as regularized least squares. This new method is available in the BioConductor R-package pengls. Inclusion of a spatial error structure improves parameter estimation and predictive model performance in low-dimensional settings and also improves feature selection in high-dimensional settings by reducing "red-shift": the preferential selection of features with spatial structure. In addition, we argue that the absence of spatial autocorrelation (SAC) in the model residuals should not be taken as a sign of a good fit, since it may result from overfitting the spatial trend. Finally, we confirm our findings in a case study on the prediction of winter wheat yield based on multispectral measurements.

Integrating Drone Technology into an Innovative Agrometeorological Methodology for the Precise and Real-Time Estimation of Crop Water Requirements

Precision agriculture has been at the cutting edge of research during the recent decade, aiming to reduce water consumption and ensure sustainability in agriculture. The proposed methodology was based on the crop water stress index (CWSI) and was applied in Greece within the ongoing research project GreenWaterDrone. The innovative approach combines real spatial data, such as infrared canopy temperature, air temperature, air relative humidity, and thermal infrared image data, taken above the crop field using an aerial micrometeorological station (AMMS) and a thermal (IR) camera installed on an unmanned aerial vehicle (UAV). Following an initial calibration phase, where the ground micrometeorological station (GMMS) was installed in the crop, no equipment needed to be maintained in the field. Aerial and ground measurements were transferred in real time to sophisticated databases and applications over existing mobile networks for further processing and estimation of the actual water requirements of a specific crop at the field level, dynamically alerting/informing local farmers/agronomists of the irrigation necessity and additionally for potential risks concerning their fields. The supported services address farmers’, agricultural scientists’, and local stakeholders’ needs to conform to regional water management and sustainable agriculture policies. As preliminary results of this study, we present indicative original illustrations and data from applying the methodology to assess UAV functionality while aiming to evaluate and standardize all system processes.

Phenomics-Assisted Selection for Herbage Accumulation in Alfalfa (Medicago sativa L.)

The application of remote sensing in plant breeding is becoming a routine method for fast and non-destructive high-throughput phenotyping (HTP) using unmanned aerial vehicles (UAVs) equipped with sensors. Alfalfa (Medicago sativa L.) is a perennial forage legume grown in more than 30 million hectares worldwide. Breeding alfalfa for herbage accumulation (HA) requires frequent and multiple phenotyping efforts, which is laborious and costly. The objective of this study was to assess the efficiency of UAV-based imagery and spatial analysis in the selection of alfalfa for HA. The alfalfa breeding population was composed of 145 full-sib and 34 half-sib families, and the experimental design was a row-column with augmented representation of controls. The experiment was established in November 2017, and HA was harvested four times between August 2018 and January 2019. A UAV equipped with a multispectral camera was used for HTP before each harvest. Four vegetation indices (VIs) were calculated from the UAV-based images: NDVI, NDRE, GNDVI, and GRVI. All VIs showed a high correlation with HA, and VIs predicted HA with moderate accuracy. HA and NDVI were used for further analyses to calculate the genetic parameters using linear mixed models. The spatial analysis had a significant effect in both dimensions (rows and columns) for HA and NDVI, resulting in improvements in the estimation of genetic parameters. Univariate models for NDVI and HA, and bivariate models, were fit to predict family performance for scenarios with various levels of HA data (simulated in silico by assigning missing values to full dataset). The bivariate models provided higher correlation among predicted values, higher coincidence for selection, and higher genetic gain even for scenarios with only 30% of HA data. Hence, HTP is a reliable and efficient method to aid alfalfa phenotyping to improve HA. Additionally, the use of spatial analysis can also improve the accuracy of selection in breeding trials.