Automatic Detection of Flavescence Dorée Symptoms Across White Grapevine Varieties Using Deep Learning

Justine Boulent, Pierre-Luc St-Charles, Samuel Foucher, Jérôme ThéauFrontiers Artif. Intell 2020 · DOI · OpenReview

Introduction

Flavescence dorée (FD) is a highly contagious and devastating phytoplasma disease that poses a significant threat to European viticulture. Currently, there is no known cure for FD, making early detection and the subsequent removal of infected vines the only viable management strategy. This "search and destroy" mission typically relies on manual scouting by trained experts, a process that is labor-intensive, costly, and susceptible to human error.

The Challenge of Automated Detection

Automating FD detection through computer vision is a complex task. The visual symptoms of the disease—including the yellowing of leaves, a lack of shoot lignification, and characteristic leaf curling—often overlap with other conditions such as nutrient deficiencies or downy mildew.

Figure 1: Visual manifestations of Flavescence dorée symptoms on grapevine leaves, showing varying degrees of yellowing and necrosis.

Furthermore, these symptoms manifest differently across various grape varieties. This phenotypic diversity makes it difficult to develop a universal detection tool that performs reliably across a diverse vineyard landscape.

Methodology

Researchers developed a two-stage deep learning pipeline designed to maximize detection accuracy while minimizing the need for extensive, pixel-level data annotation:

  1. Patch-Based Classification: A ResNet18 Convolutional Neural Network (CNN) was pre-trained on ImageNet and fine-tuned using small image patches of Chardonnay leaves. This approach allowed the model to learn specific features of diseased vs. healthy tissue from a relatively small dataset of approximately 7,300 annotated patches.
  2. FCN Conversion for Field Inference: To facilitate practical application, the trained classifier was converted into a Fully Convolutional Network (FCN). This architectural shift enables the model to process full-sized field images and generate coarse "heat maps" indicating symptomatic areas, providing localization without requiring expensive pixel-perfect manual segmentation for training.
  3. Explainability and Visualization: The team utilized Guided Grad-CAM to visualize the specific features (such as leaf veins or curls) that triggered model predictions. Additionally, UMAP (Uniform Manifold Approximation and Projection) was used to analyze how the model clustered different varieties within its internal feature space.

Experimental Setup and Datasets

The study evaluated the model's robustness across several datasets:

  • Training Set: Focused primarily on Chardonnay vines.
  • Personal Test Sets: 153 Chardonnay and 155 Ugni-Blanc images captured by the research team.
  • External Test Sets: 34 images from FREDON Aquitaine covering six distinct varieties: Chardonnay, Ugni-Blanc, Merlot Blanc, Riesling, Exalta, and Semillon. This served as a critical "blind" test for real-world generalization.
Figure 2: Comparison of Flavescence dorée symptoms across six white grapevine varieties: Chardonnay, Semillon, Exalta, Merlot Blanc, Riesling, and Ugni-Blanc.

Key Results and the "Varietal Wall"

The model demonstrated exceptional performance on its training variety, achieving a 98.48% True Positive Rate (TPR) on the personal Chardonnay dataset and 100% on the external Chardonnay samples. The False Positive Rate (FPR) remained impressively low at approximately 1.15%, suggesting the tool is unlikely to cause unnecessary vine uprooting.

However, a significant "varietal gap" emerged during cross-variety testing. While the model maintained high accuracy on varieties with similar symptomatic curling (e.g., Merlot Blanc and Riesling), its performance plummeted to just 8.3% TPR for Ugni-Blanc. Ugni-Blanc typically exhibits yellowing without the pronounced curling found in Chardonnay, a feature the model had heavily weighted during training.

Discussion and Explainability

Explainability tools provided vital insights into these failures. Guided Grad-CAM revealed that the model occasionally focused on irrelevant environmental features, such as trellising wires or shadows, rather than biological symptoms. UMAP visualizations further confirmed that while healthy vines of different varieties share similar feature spaces, their diseased states diverge significantly. This divergence explains why a model trained on a single variety struggles to generalize.

Conclusion

This research highlights both the potential and the current limitations of deep learning in viticulture. While the pipeline is highly effective for Chardonnay, it currently functions as a "Chardonnay specialist." To transition from a research prototype to a commercial-grade tool, future work must incorporate multi-varietal, multi-year datasets to account for the diverse ways FD manifests across different climates and grape varieties. Implementing "active learning" strategies may further help the model efficiently expand its expertise across the global viticultural landscape.

Summary updated

Abstract

Flavescence dorée (FD) is a grapevine disease caused by phytoplasmas and transmitted by leafhoppers that has been spreading in European vineyards despite significant efforts to control it. In this study, we aim to develop a model for the automatic detection of FD-like symptoms (which encompass other grapevine yellows symptoms)}. The concept is to detect likely FD-affected grapevines so that samples can be removed for FD laboratory identification, followed by uprooting if they test positive, all to be conducted quickly and without omission, thus avoiding further contamination in the fields. Developing FD-like symptoms detection models is not simple, as it requires dealing with the complexity of field conditions and FD symptoms' expression. To address these challenges, we use deep learning, which has already been proven effective in similar contexts. More specifically, we train a Convolutional Neural Network (CNN) on image patches, and convert it into a Fully Convolutional Network (FCN) to perform inference. As a result, we obtain a coarse segmentation of the likely FD-affected areas while having only trained a classifier, which is less demanding in terms of annotations. We evaluate the performance of our model trained on a white grape variety, Chardonnay, across five other grape varieties with varying FD symptoms expressions. Of the two largest test datasets, the true positive rate for Chardonnay reaches 98.48\% whereas for Ugni-Blanc it drops to 8.3\%, underlining the need for a multi-varietal training dataset to capture the diversity of FD symptoms. To obtain more transparent results and to better understand the model's sensitivity, we investigate its behavior using two visualization techniques, Guided Gradient-weighted Class Activation Mapping (GG-CAM) and the Uniform Manifold Approximation and Projection (UMAP). Such techniques lead to a more comprehensive analysis with greater reliability, which is essential for in-field applications, and more broadly, for all applications impacting humans and the environment.
BibTeX
@article{boulent2020automatic,
  author = {Justine Boulent and Pierre-Luc St-Charles and Samuel Foucher and Jérôme Théau},
  title = {{Automatic Detection of Flavescence Dorée Symptoms Across White Grapevine Varieties Using Deep Learning}},
  journal = {Frontiers Artif. Intell.},
  year = {2020},
  doi = {10.3389/frai.2020.564878},
}
Figure