terclim by ICS banner
IVES 9 IVES Conference Series 9 GiESCO 9 Implementation of a deep learning-based approach for detecting and localising automatically grapevine leaves with downy mildew symptoms

Implementation of a deep learning-based approach for detecting and localising automatically grapevine leaves with downy mildew symptoms

Abstract

Context and purpose of the study – Grapevine downy mildew is a disease of foliage caused by Oomycete Plasmopara viticola an endoparasite that develops inside grapevine organs and can infect virtually every green organ. Downy mildew is one of the most destructive diseases in wine-growing regions, drastically reducing yield and fruit quality. Traditional manual disease detection relies on farm experts. Human field scouting has been widely used for monitoring the disease progress, however, is costly, laborious, subjective, and often imprecise. In this sense, computer vision technologies and artificial intelligence provide a suitable alternative to improve the current disease detection techniques. Therefore, this study aims to validate a deep learning-based approach for detecting and localising automatically leaves with downy mildew symptoms.

Material and methods – Fourteen commercial blocks (different grapevine cultivars) located in northern Spain were assessed to generate a comprehensive dataset for validating the deep learning algorithm. All analysed blocks presented downy mildew symptoms with different levels of intensity. RGB Images of the canopy were taken manually using a conventional camera. The images were acquired during different daily hours and under contrasting light conditions. YOLOv4 (You Only Look Once) was the deep learning algorithm analysed in this study. YOLOv4 model was trained using a heterogeneous dataset populated by RGB images obtained from the different vineyard blocks under different conditions to increase the robustness of the model. The RGB images were carefully labelled manually by an expert, selecting leaves with visible downy mildew symptoms. The labelled images were divided into two datasets, using 80% for training and validating and 20% for testing. The images used for training were divided into 1500 x 1500 pixel sub-images obtaining 15 sub-images per image and the sub-images were resized to 640×640 pixels. Data processing and deep learning modelling were performed with Python programming language and the Darknet neural network framework. The metrics used to evaluate the model were mean Average Precision (mAP), F1-score and Intersection over Union (IoU).

Results – The results of YOLOv4 for detecting leaves with downy mildew symptoms are promising. In the testing process applied on full images (2560×1728 pixels), the model presented a mAP of 67%, an F1-score of 0.69 and an IoU of 62%. When the number of real infected leaves (labelled by an expert) was compared with the predicted number of infected leaves the model reached a determination coefficient R2 of 0.93.  The accuracy of the method to determine the number of infected leaves was similar across the whole range of infections. This indicates that the model fits appropriately to all conditions tested. Also, the analysis of the localization indicates that leaves with apparent symptoms were detected correctly by the model.

DOI:

Publication date: July 5, 2023

Issue: GiESCO 2023

Type: Poster

Authors

Carlos POBLETE-ECHEVERRÍA1,2, Inés HERNÁNDEZ1,2, Salvador GUTIÉRREZ3, Rubén ÍÑIGUEZ1,2, Ignacio BARRIO1,2 and Javier TARDAGUILA1,2*

1Televitis Research Group, University of La Rioja, 26006 Logroño, Spain
2Institute of Grapevine and Wine Sciences (University of La Rioja, Consejo Superior de Investigaciones Científicas, Gobierno de La Rioja), 26007 Logroño, Spain
3Department of Computer Science and Artificial Intelligence (DECSAI), Andalusian Research Institute in Data Science and Computational Intelligence (DaSCI), University of Granada (UGR), 18071 Granada, Spain

Contact the author*

Keywords

artificial intelligence, deep learning, disease detection, precision viticulture, YOLOv4

Tags

GiESCO | GIESCO 2023 | IVES Conference Series

Citation

Related articles…

Climats: a model of terroir-based winegrowing recognized by UNESCO

In Burgundy, a climat has nothing to do with the weather but accurately designates a named vine plot, often centuries-old, which produces a singular wine. This wine is the combination of history, the natural environment (relief, type of soil, exposure to the sun), a grape variety and know-how going back thousands of years. The grapes of each climat are harvested separately and the wine is made from a single grape variety and has a unique name featured on the bottle. Romanée conti, clos de vougeot, montrachet, musigny, corton…

Lean management to improve sustainability in wine sector: an exploratory study in the Prosecco DOC appellation

The contemporary wine sector confronts a formidable array of challenges, including burgeoning production costs and the constricted availability of natural resources. Heightened consumer awareness regarding sustainability issues further compounds these pressures, compelling companies to adopt more judicious resource utilization strategies. In response to these imperatives, there is a growing recognition of the need to overhaul production methodologies within the wine industry with a view to minimizing inputs and eliminating waste.

Emosensory profile and chemical characterization of wine vinegar from the Douro and Rioja demarcated regions

Wine vinegars have a tangy flavor and are versatile in cooking. They’ve been used since the neolithic period and are now used as microbial inhibitors and acidifiers. They’re low in calories, have antioxidants, and have a long shelf life, but quality may decrease after opening. The objective of this study focuses on the physical-chemical, sensory, and emotional characterization of wine vinegar samples from the douro demarcated region and la rioja. In total, 22 samples of wine vinegar were analyzed at the time of opening.

Where the sky is no limit — The transformation of wine marketing through text-to-video generation AI model

The introduction of ai-driven tools in digital content creation represents a significant shift in the landscape of marketing, particularly for industries reliant on rich visual storytelling such as the wine sector. The development of ai models like openai’s sora, runway’s gen-2 or google’s lumiere, which can generate realistic video content from textual descriptions, offers promising new avenues for enhancing brand narrative and consumer engagement. This research explores the potential of text-to-video (t2v) ai models to revolutionize wine marketing by creating dynamic, engaging content that captures the essence of vineyards and their products without the need for traditional video production processes.

100 ans d’évolution des règles relatives à l’encépagement des AOC viticoles françaises : quelles perspectives face aux enjeux contemporains

To characterize a wine, the most frequently used criteria describe its color, its origin, the grape varieties from which they come, or even for white wines its residual sugar content (dry, semi-dry, sweet). In france, the system of appellations of origin set up in 1919 was initially based solely on the notoriety and origin of the wines. But given the unfavorable consequences that this lack of details generated, the public authorities quickly integrated in 1927 into the “capus” law criteria for access to designations of origin, relating to the specific characteristics of the soils of the vineyards and the grape varieties used, in particular exclusion of interspecific hybrid varieties. In 1935 the creation of the aoc system confirmed the interest in precisely defining all the production conditions that must be implemented to be able to claim the benefit of an aoc, and grape varieties were an essential condition for acquisition.