Journal: Plant Phenomics

Abbreviation

Publisher

Elsevier

Journal Volumes

ISSN

2643-6515

Description

Filters

20252
Reset filters

Search Results

Publications 1 - 2 of 2
  • The Global Wheat Full Semantic Organ Segmentation (GWFSS) dataset
    Item type: Journal Article
    Wang, Zijian; Zenkl, Radek Zenkl; Greche, Latifa; et al. (2025)
    Plant Phenomics
    Computer vision is increasingly used in farmers' fields and agricultural experiments to quantify important traits. Imaging setups with a sub-millimeter ground sampling distance enable the detection and tracking of plant features, including size, shape, and colour. Although today's AI-driven foundation models segment almost any object in an image, they still fail for complex plant canopies. To improve model performance, the global wheat dataset consortium assembled a diverse set of images from experiments around the globe. After the head detection dataset (GWHD), the new dataset targets a full semantic segmentation (GWFSS) of organs (leaves, stems and spikes) covering all developmental stages. Images were collected by 11 institutions using a wide range of imaging setups. Two datasets are provided: i) a set of 1096 diverse images in which all organs were labelled at the pixel level, and (ii) a dataset of 52,078 images without annotations available for additional training. The labelled set was used to train segmentation models based on DeepLabV3Plus and Segformer. Our Segformer model performed slightly better than DeepLabV3Plus with a mIOU for leaves and spikes of ca. 90 ​%. However, the precision for stems with 54 ​% was rather lower. The major advantages over published models are: i) the exclusion of weeds from the wheat canopy, ii) the detection of all wheat features including necrotic and senescent tissues and its separation from crop residues. This facilitates further development in classifying healthy vs. unhealthy tissue to address the increasing need for accurate quantification of senescence and diseases in wheat canopies.
  • Analysis of variance and its sources in UAV-based multi-view thermal imaging of wheat plots
    Item type: Journal Article
    Treier, Simon; Roth, Lukas; Hund, Andreas; et al. (2025)
    Plant Phenomics
    Canopy temperature (CT) estimates from drone-based uncooled thermal cameras are prone to confounding effects, which affects the interpretability of CT estimates. Experimental sources of variance, such as genotypes and experimental treatments blend with confounding sources of variance such as thermal drift, spatial field trends, and effects related to viewing geometry. Nevertheless, CT is gaining popularity to characterize crop performance and crop water use, and as a proxy measurement of stomatal conductance and transpiration. Drone-based thermography was therefore proposed to measure CT in agricultural experiments. For a meaningful interpretation of CT, confounding sources of variance must be considered. In this study, the multi-view approach was applied to examine the variance components of CT on 99 flights with a drone-based thermal camera. Flights were conducted on two variety testing field trials of winter wheat over two years with contrasting meteorological conditions in the temperate climate of Switzerland. It was demonstrated how experimental sources of variance can be disentangled from confounding sources of variance and on average more than 96.5 % of the initial variance could be explained with experimental and confounding sources combined. Not considering confounding sources led to erroneous conclusions about phenotypic correlations of CT with traits such as yield, plant height, fractional canopy cover, and multispectral indices. Based on extensive and diverse data, this study provides comprehensive insights into the manifold sources of variance in CT measurements, which supports the planning and interpretation of drone-based CT screenings in variety testing, breeding, and research.
Publications 1 - 2 of 2