Detection of maize stem diameter by using RGB-D cameras' depth information under selected field condition

Stem diameter is a critical phenotypic parameter for maize, integral to yield prediction and lodging resistance assessment. Traditionally, the quantification of this parameter through manual measurement has been the norm, notwithstanding its tedious and laborious nature. To address these challenges, this study introduces a non-invasive field-based system utilizing depth information from RGB-D cameras to measure maize stem diameter. This technology offers a practical solution for conducting rapid and non-destructive phenotyping. Firstly, RGB images, depth images, and 3D point clouds of maize stems were captured using an RGB-D camera, and precise alignment between the RGB and depth images was achieved. Subsequently, the contours of maize stems were delineated using 2D image processing techniques, followed by the extraction of the stem's skeletal structure employing a thinning-based skeletonization algorithm. Furthermore, within the areas of interest on the maize stems, horizontal lines were constructed using points on the skeletal structure, resulting in 2D pixel coordinates at the intersections of these horizontal lines with the maize stem contours. Subsequently, a back-projection transformation from 2D pixel coordinates to 3D world coordinates was achieved by combining the depth data with the camera's intrinsic parameters. The 3D world coordinates were then precisely mapped onto the 3D point cloud using rigid transformation techniques. Finally, the maize stem diameter was sensed and determined by calculating the Euclidean distance between pairs of 3D world coordinate points. The method demonstrated a Mean Absolute Percentage Error (MAPE) of 3.01%, a Mean Absolute Error (MAE) of 0.75 mm, a Root Mean Square Error (RMSE) of 1.07 mm, and a coefficient of determination (R²) of 0.96, ensuring accurate measurement of maize stem diameter. This research not only provides a new method of precise and efficient crop phenotypic analysis but also offers theoretical knowledge for the advancement of precision agriculture.

Overcoming field variability: unsupervised domain adaptation for enhanced crop-weed recognition in diverse farmlands

Recent developments in deep learning-based automatic weeding systems have shown promise for unmanned weed eradication. However, accurately distinguishing between crops and weeds in varying field conditions remains a challenge for these systems, as performance deteriorates when applied to new or different fields due to insignificant changes in low-level statistics and a significant gap between training and test data distributions. In this study, we propose an approach based on unsupervised domain adaptation to improve crop-weed recognition in new, unseen fields. Our system addresses this issue by learning to ignore insignificant changes in low-level statistics that cause a decline in performance when applied to new data. The proposed network includes a segmentation module that produces segmentation maps using labeled (training field) data while also minimizing entropy using unlabeled (test field) data simultaneously, and a discriminator module that maximizes the confusion between extracted features from the training and test farm samples. This module uses adversarial optimization to make the segmentation network invariant to changes in the field environment. We evaluated the proposed approach on four different unseen (test) fields and found consistent improvements in performance. These results suggest that the proposed approach can effectively handle changes in new field environments during real field inference.

Functional phenomics for improved climate resilience in Nordic agriculture

The five Nordic countries span the most northern region for field cultivation in the world. This presents challenges per se, with short growing seasons, long days, and a need for frost tolerance. Climate change has additionally increased risks for micro-droughts and water logging, as well as pathogens and pests expanding northwards. Thus, Nordic agriculture demands crops that are adapted to the specific Nordic growth conditions and future climate scenarios. A focus on crop varieties and traits important to Nordic agriculture, including the unique resource of nutritious wild crops, can meet these needs. In fact, with a future longer growing season due to climate change, the region could contribute proportionally more to global agricultural production. This also applies to other northern regions, including the Arctic. To address current growth conditions, mitigate impacts of climate change, and meet market demands, the adaptive capacity of crops that both perform well in northern latitudes and are more climate resilient has to be increased, and better crop management systems need to be built. This requires functional phenomics approaches that integrate versatile high-throughput phenotyping, physiology, and bioinformatics. This review stresses key target traits, the opportunities of latitudinal studies, and infrastructure needs for phenotyping to support Nordic agriculture.

Crop phenotyping in a context of global change: What to measure and how to do it

High-throughput crop phenotyping, particularly under field conditions, is nowadays perceived as a key factor limiting crop genetic advance. Phenotyping not only facilitates conventional breeding, but it is necessary to fully exploit the capabilities of molecular breeding, and it can be exploited to predict breeding targets for the years ahead at the regional level through more advanced simulation models and decision support systems. In terms of phenotyping, it is necessary to determined which selection traits are relevant in each situation, and which phenotyping tools/methods are available to assess such traits. Remote sensing methodologies are currently the most popular approaches, even when lab-based analyses are still relevant in many circumstances. On top of that, data processing and automation, together with machine learning/deep learning are contributing to the wide range of applications for phenotyping. This review addresses spectral and red-green-blue sensing as the most popular remote sensing approaches, alongside stable isotope composition as an example of a lab-based tool, and root phenotyping, which represents one of the frontiers for field phenotyping. Further, we consider the two most promising forms of aerial platforms (unmanned aerial vehicle and satellites) and some of the emerging data-processing techniques. The review includes three Boxes that examine specific case studies.

Advances in Cereal Crop Genomics for Resilience under Climate Change

Adapting to climate change, providing sufficient human food and nutritional needs, and securing sufficient energy supplies will call for a radical transformation from the current conventional adaptation approaches to more broad-based and transformative alternatives. This entails diversifying the agricultural system and boosting productivity of major cereal crops through development of climate-resilient cultivars that can sustainably maintain higher yields under climate change conditions, expanding our focus to crop wild relatives, and better exploitation of underutilized crop species. This is facilitated by the recent developments in plant genomics, such as advances in genome sequencing, assembly, and annotation, as well as gene editing technologies, which have increased the availability of high-quality reference genomes for various model and non-model plant species. This has necessitated genomics-assisted breeding of crops, including underutilized species, consequently broadening genetic variation of the available germplasm; improving the discovery of novel alleles controlling important agronomic traits; and enhancing creation of new crop cultivars with improved tolerance to biotic and abiotic stresses and superior nutritive quality. Here, therefore, we summarize these recent developments in plant genomics and their application, with particular reference to cereal crops (including underutilized species). Particularly, we discuss genome sequencing approaches, quantitative trait loci (QTL) mapping and genome-wide association (GWAS) studies, directed mutagenesis, plant non-coding RNAs, precise gene editing technologies such as CRISPR-Cas9, and complementation of crop genotyping by crop phenotyping. We then conclude by providing an outlook that, as we step into the future, high-throughput phenotyping, pan-genomics, transposable elements analysis, and machine learning hold much promise for crop improvements related to climate resilience and nutritional superiority.

Potential of Hyperspectral and Thermal Proximal Sensing for Estimating Growth Performance and Yield of Soybean Exposed to Different Drip Irrigation Regimes Under Arid Conditions

Proximal hyperspectral sensing tools could complement and perhaps replace destructive traditional methods for accurate estimation and monitoring of various morpho-physiological plant indicators. In this study, we assessed the potential of thermal imaging (TI) criteria and spectral reflectance indices (SRIs) to monitor different vegetative growth traits (biomass fresh weight, biomass dry weight, and canopy water mass) and seed yield (SY) of soybean exposed to 100%, 75%, and 50% of estimated crop evapotranspiration (ETc). These different plant traits were evaluated and related to TI criteria and SRIs at the beginning bloom (R1) and full seed (R6) growth stages. Results showed that all plant traits, TI criteria, and SRIs presented significant variations (p < 0.05) among irrigation regimes at both growth stages. The performance of TI criteria and SRIs for assessment of vegetative growth traits and SY fluctuated when relationships were analyzed for each irrigation regime or growth stage separately or when the data of both conditions were combined together. TI criteria and SRIs exhibited a moderate to strong relationship with vegetative growth traits when data from different irrigation regimes were pooled together at each growth stage or vice versa. The R6 and R1 growth stages are suitable for assessing SY under full (100% ETc) and severe (50% ETc) irrigation regimes, respectively, using SRIs. The overall results indicate that the usefulness of the TI and SRIs for assessment of growth, yield, and water status of soybean under arid conditions is limited to the growth stage, the irrigation level, and the combination between them.

Imaging Wheat Canopy Through Stereo Vision: Overcoming the Challenges of the Laboratory to Field Transition for Morphological Features Extraction

Stereo vision is a 3D imaging method that allows quick measurement of plant architecture. Historically, the method has mainly been developed in controlled conditions. This study identified several challenges to adapt the method to natural field conditions and propose solutions. The plant traits studied were leaf area, mean leaf angle, leaf angle distribution, and canopy height. The experiment took place in a winter wheat, Triticum aestivum L., field dedicated to fertilization trials at Gembloux (Belgium). Images were acquired thanks to two nadir cameras. A machine learning algorithm using RGB and HSV color spaces is proposed to perform soil-plant segmentation robust to light conditions. The matching between images of the two cameras and the leaf area computation was improved if the number of pixels in the image of a scene was binned from 2560 × 2048 to 1280 × 1024 pixels, for a distance of 1 m between the cameras and the canopy. Height descriptors such as median or 95th percentile of plant heights were useful to precisely compare the development of different canopies. Mean spike top height was measured with an accuracy of 97.1 %. The measurement of leaf area was affected by overlaps between leaves so that a calibration curve was necessary. The leaf area estimation presented a root mean square error (RMSE) of 0.37. The impact of wind on the variability of leaf area measurement was inferior to 3% except at the stem elongation stage. Mean leaf angles ranging from 53° to 62° were computed for the whole growing season. For each acquisition date during the vegetative stages, the variability of mean angle measurement was inferior to 1.5% which underpins that the method is precise.

Novel and Automatic Rice Thickness Extraction Based on Photogrammetry Using Rice Edge Features

The dimensions of phenotyping parameters such as the thickness of rice play an important role in rice quality assessment and phenotyping research. The objective of this study was to propose an automatic method for extracting rice thickness. This method was based on the principle of binocular stereovision but avoiding the problem that it was difficult to directly match the corresponding points for 3D reconstruction due to the lack of texture of rice. Firstly, the shape features of edge, instead of texture, was used to match the corresponding points of the rice edge. Secondly, the height of the rice edge was obtained by way of space intersection. Finally, the thickness of rice was extracted based on the assumption that the average height of the edges of multiple rice is half of the thickness of rice. According to the results of the experiments on six kinds of rice or grain, errors of thickness extraction were no more than the upper limit of 0.1 mm specified in the national industry standard. The results proved that edge features could be used to extract rice thickness and validated the effectiveness of the thickness extraction algorithm we proposed, which provided technical support for the extraction of phenotyping parameters for crop researchers.

High-Throughput Phenotyping of Canopy Cover and Senescence in Maize Field Trials Using Aerial Digital Canopy Imaging

In the crop breeding process, the use of data collection methods that allow reliable assessment of crop adaptation traits, faster and cheaper than those currently in use, can significantly improve resource use efficiency by reducing selection cost and can contribute to increased genetic gain through improved selection efficiency. Current methods to estimate crop growth (ground canopy cover) and leaf senescence are essentially manual and/or by visual scoring, and are therefore often subjective, time consuming, and expensive. Aerial sensing technologies offer radically new perspectives for assessing these traits at low cost, faster, and in a more objective manner. We report the use of an unmanned aerial vehicle (UAV) equipped with an RGB camera for crop cover and canopy senescence assessment in maize field trials. Aerial-imaging-derived data showed a moderately high heritability for both traits with a significant genetic correlation with grain yield. In addition, in some cases, the correlation between the visual assessment (prone to subjectivity) of crop senescence and the senescence index, calculated from aerial imaging data, was significant. We concluded that the UAV-based aerial sensing platforms have great potential for monitoring the dynamics of crop canopy characteristics like crop vigor through ground canopy cover and canopy senescence in breeding trial plots. This is anticipated to assist in improving selection efficiency through higher accuracy and precision, as well as reduced time and cost of data collection.

Characterising root trait variability in chickpea (Cicer arietinum L.) germplasm

Chickpea (Cicer arietinum L.) is an important grain legume crop but its sustainable production is challenged by predicted climate changes, which are likely to increase production limitations and uncertainty in yields. Characterising the variability in root architectural traits in a core collection of chickpea germplasm will provide the basis for breeding new germplasm with suitable root traits for the efficient acquisition of soil resources and adaptation to drought and other abiotic stresses. This study used a semi-hydroponic phenotyping system for assessing root trait variability across 270 chickpea genotypes. The genotypes exhibited large variation in rooting patterns and branching manner. Thirty root-related traits were characterised, 17 of which had coefficients of variation ≥0.3 among genotypes and were selected for further examination. The Pearson correlation matrix showed a strong correlation among most of the selected traits (P≤0.05). Principal component analysis revealed three principal components with eigenvalues >1 capturing 81.5% of the total variation. An agglomerative hierarchical clustering analysis, based on root trait variation, identified three genotype homogeneous groups (rescaled distance of 15) and 16 sub-groups (rescaled distance of 5). The chickpea genotypes characterised in this study with vastly different root properties could be used for further studies in glasshouses and field trials, and for molecular marker studies, gene mapping, and modelling simulations, ultimately aimed at breeding germplasm with root traits for improved adaptation to drought and other specific environments.