The detection of wheat heads in plant images is an important task for estimating pertinent wheat traits including head population density and head characteristics such as health, size, maturity stage, and the presence of awns. Several studies have developed methods for wheat head detection from high-resolution RGB imagery based on machine learning algorithms. However, these methods have generally been calibrated and validated on limited datasets. High variability in observational conditions, genotypic differences, development stages, and head orientation makes wheat head detection a challenge for computer vision. Further, possible blurring due to motion or wind and overlap between heads for dense populations make this task even more complex. Through a joint international collaborative effort, we have built a large, diverse, and well-labelled dataset of wheat images, called the Global Wheat Head Detection (GWHD) dataset. It contains 4700 high-resolution RGB images and 190000 labelled wheat heads collected from several countries around the world at different growth stages with a wide range of genotypes. Guidelines for image acquisition, associating minimum metadata to respect FAIR principles, and consistent head labelling methods are proposed when developing new head detection datasets. The GWHD dataset is publicly available at http://www.global-wheat.com/and aimed at developing and benchmarking methods for wheat head detection.
The Global Wheat Head Detection (GWHD) dataset was created in 2020 and has assembled 193,634 labelled wheat heads from 4700 RGB images acquired from various acquisition platforms and 7 countries/institutions. With an associated competition hosted in Kaggle, GWHD_2020 has successfully attracted attention from both the computer vision and agricultural science communities. From this first experience, a few avenues for improvements have been identified regarding data size, head diversity, and label reliability. To address these issues, the 2020 dataset has been reexamined, relabeled, and complemented by adding 1722 images from 5 additional countries, allowing for 81,553 additional wheat heads. We now release in 2021 a new version of the Global Wheat Head Detection dataset, which is bigger, more diverse, and less noisy than the GWHD_2020 version.
26 February 2021 in Soil Science Society of America Journal
Soil compaction by farm machinery may persist for decades, hampering soil productivity and functioning. Assessing compaction costs and guiding recovery strategies are hindered by paucity of data on soil structure recovery rates. A long‐term Soil Structure Observatory was established on a loamy soil in Switzerland to monitor soil structure recovery after prescribed compaction, and to better assess the roles of natural processes (vegetation, macrofauna, and shrink–swell cycles) on recovery patterns. The aim of this study was to quantify short‐term soil structure recovery under natural conditions in the presence and absence of plant cover (ley and bare soil). We measured soil porosity and gas and water transport capabilities at 0.1 and 0.3 m depth. Two years after the compaction event, soil physical properties have not recovered to precompaction levels, even within the topsoil. Surprisingly, no differences were observed in the recovery patterns of ley and bare soil treatments. Measurements show that recovery rates differ among soil properties with the most severely affected properties by compaction (permeability) exhibiting highest recovery rates. Total soil porosity shows no recovery trend, suggesting lack of soil decompaction. Improved soil functions and decompaction are distinct aspects of soil structure recovery, with the latter requiring net upward transport of soil mass. We suggest that soil structure recovery proceeds at two fronts: from the soil surface downward, and expanding around local biologically‐active pockets (marked by biopores) into the compacted soil volumes. This concept could be tested with additional data of longer time series at our site as well as in other soils and climates.
Early generation breeding nurseries with thousands of genotypes in single-row plots are well suited to capitalize on high throughput phenotyping. Nevertheless, methods to monitor the intrinsically hard-to-phenotype early development of wheat are yet rare. We aimed to develop proxy measures for the rate of plant emergence, the number of tillers, and the beginning of stem elongation using drone-based imagery. We used RGB images (ground sampling distance of 3 mm pixel-1) acquired by repeated flights (≥ 2 flights per week) to quantify temporal changes of visible leaf area. To exploit the information contained in the multitude of viewing angles within the RGB images, we processed them to multiview ground cover images showing plant pixel fractions. Based on these images, we trained a support vector machine for the beginning of stem elongation (GS30). Using the GS30 as key point, we subsequently extracted plant and tiller counts using a watershed algorithm and growth modeling, respectively. Our results show that determination coefficients of predictions are moderate for plant count (), but strong for tiller count () and GS30 (). Heritabilities are superior to manual measurements for plant count and tiller count, but inferior for GS30 measurements. Increasing the selection intensity due to throughput may overcome this limitation. Multiview image traits can replace hand measurements with high efficiency (85–223%). We therefore conclude that multiview images have a high potential to become a standard tool in plant phenomics.
Predicting plant development, a longstanding goal in plant physiology, involves 2 interwoven components: continuous growth and the progression of growth stages (phenology). Current models for winter wheat and soybean assume species-level growth responses to temperature. We challenge this assumption, suggesting that cultivar-specific temperature responses substantially affect phenology. To investigate, we collected field-based growth and phenology data in winter wheat and soybean over multiple years. We used diverse models, from linear to neural networks, to assess growth responses to temperature at various trait and covariate levels. Cultivar-specific nonlinear models best explained phenology-related cultivar–environment interactions. With cultivar-specific models, additional relations to other stressors than temperature were found. The availability of the presented field phenotyping tools allows incorporating cultivar-specific temperature response functions in future plant physiology studies, which will deepen our understanding of key factors that influence plant development. Consequently, this work has implications for crop breeding and cultivation under adverse climatic conditions.
1 August 2021 in Environmental and Experimental Botany