terclim by ICS banner
IVES 9 IVES Conference Series 9 A novel dataset and deep learning object detection benchmark for grapevine pest surveillance

A novel dataset and deep learning object detection benchmark for grapevine pest surveillance

Abstract

Flavescence dorée (FD) stands out as a significant grapevine disease with severe implications for vineyards. The American grapevine leafhopper (Scaphoideus titanus) serves as the primary vector, transmitting the pathogen that causes yield losses and elevated costs linked to uprooting and replanting. Another potential vector of FD is the mosaic leafhopper, Orientus ishidae, commonly found in agroecosystems. The current monitoring approach involves periodic human identification of chromotropic traps, a labor-intensive and time-consuming process.

Therefore, there is a compelling need to develop an automatic pest detection system, leveraging the recent progress in computer vision and deep learning techniques. However, the current progress in developing such a system is hindered by the lack of effective datasets to serve as ground-truth data for the training process.

To fill this gap, our study contributes a fully annotated dataset of S. titanus and Or. ishidae from yellow sticky traps. The dataset comprises more than 400 images, with 1000 identification per class. Guided by entomologists, the annotation task involved defining bounding boxes around relevant insects with corresponding class labels.

We trained and compared the performance of state-of-the-art object detection algorithms (YOLOv8 and Faster R-CNN). Pre-processing included automatic cropping to eliminate irrelevant background information and image enhancements to improve overall quality. Additionally, we tested the impact of altering image resolution, data augmentation, and single-class detection. Preliminary results achieved a high detection accuracy, with mAP@50 and F1-score above 90%, and mAP@50-95 around 70%, allowing a first deployment as an automatic annotation support tool.

DOI:

Publication date: June 14, 2024

Issue: Open GPB 2024

Type: Poster

Authors

Giorgio Checola1*, Paolo Sonego1, Valerio Mazzoni2, Franca Ghidoni3, Alberto Gelmetti3, Pietro Franceschi1

1 Research and Innovation Centre, Digital Agriculture Unit, Fondazione Edmund Mach, S. Michele all’Adige, TN, Italy
2 Research and Innovation Centre, Plant Protection Unit, Fondazione Edmund Mach, S. Michele all’Adige, TN, Italy
3 Technology Transfer Centre, Viticulture Unit, Fondazione Edmund Mach, S. Michele all’Adige, TN, Italy

Contact the author*

Keywords

insect detection, deep learning, smart pest monitoring, flavescence dorée, insect traps

Tags

IVES Conference Series | Open GPB | Open GPB 2024

Citation

Related articles…

Evaluation of colloidal stability in white and rosé wines investing Dynamic Light Scattering technology

Proteins constitute one of the three main components of grape juice and white wine, phenolic compounds and polysaccharides being the others. A specific group of the total grape-derived proteins resists degradation or adsorption during the winemaking process and remains in finished wine if not removed by the commonplace commercial practice of bentonite fining. While bentonite is effective in removing the problematic proteins, it is claimed to adversely affect the quality of the treated wine under certain conditions, through the removal of colour, flavor and texture compounds. A number of studies have indicated that different protein fractions require distinct bentonite concentrations for protein removal and consequent heat stabilization.

REDWINE project: use of Chlorella vulgaris to prevent biotic and abiotic stress in Palmela’s region, Portugal, vineyards

The new EU Green Deal aims to achieve GHG emissions reduction by at least 55% by 2030 and a climate neutral EU economy by 2050.
REDWine concept will be realized through the establishment of an integrated Living Lab demonstrating the viability of the system at TRL 7. The Living Lab will be able to utilize 2 ton of fermentation off-gas/year (90% of total CO2 produced in the fermenter) and 80 m3 of liquid effluent (100% of the liquid effluent generated during fermenter washing) to produce 1 ton (dry weight) of Chlorella biomass/year. This biomass will be processed under a downstream extraction process to obtain added-value extracts and applied in food, cosmetic and agricultural end-products and to generate a new EcoWine. REDWine will focus on the recovery of off-gas from a 20.000L fermenter of red wine production existing in Adega Cooperativa de Palmela (ACP, located in Palmela, Portugal).

OENOLOGICAL STRATEGIES FOR THE REMOVAL OF PINKING IN WHITE WINE

The pinking of in white wine is the turning of color from yellow to salmon hue. White wines obtained from certain grape varieties (e.g. Chardonnay, Sauvignon blanc, Riesling, Trebbiano di Lugana) showed to be susceptible to pinking [1] that has been evaluated by an assay providing the addition of hydrogen peroxide. Even if its appearance does not seem to affect the sensory properties [2], strategies are necessary for its removal. Nowadays, the treatment with polyvinylpolipirroline (PVPP) was reported to significantly decrease the pink color [3].

Chitosan treatment to manage grapevine downy mildew

Downy mildew is one of the most important grapevine diseases, caused by the Oomycete Plasmopara viticola. The management of the disease in organic agriculture can require up to 15 copper applications per year. However, copper accumulates in the soil, is phytotoxic and is toxic for organisms living in the soil, its use has been restricted in European Union to maximum 28 kg in 7 years. Therefore, testing of alternatives with equal effectiveness is desirable. Among those, the natural biopolymer chitosan, obtained from crab shells, proved to be effective toward downy mildew in plot experiments. The aim of our trials was to extend chitosan applications in large scale experiments in different years, cultivars and environmental conditions.

Zonage et caractérisation des terroirs de l’AOC Côtes-du-Rhône: exemple du bassin de Nyons-Valreas

The southern Côtes-du-Rhône vineyard shows a significant variety of ecological facets over the Lower Rhone Valley. Intending to characterize such a variety of “terroir “called vineyard situations, a spatial approach based on identification of soil landscapes has been initiated.