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…

OPTIMIZATION, VALIDATION AND APPLICATION OF THE EPR SPIN-TRAPPING TECHNIQUE TO THE DETECTION OF FREE RADICALS IN CHARDONNAY WINES

The aging potential of Burgundy chardonnay wines is considered as quality indicator. However, some of them exhibit higher oxidative sensitivity and premature oxidative aging symptoms, which are potentially induced by no-enzymatic oxidation such as Fenton-type reaction (Danilewicz, 2003). This chemical mechanism involves the action of transition metal, native phenolic compounds and oxygen which promote the generation of highly reactive oxygen species (ROS) such as hydroxyl radicals (OH) or 1-hydroxyethyl radicals (1-HER) from oxidation of ethanol. Such mechanism is involved in the radical oxidation occurring during bottle aging. According to Elias et al.,(2009a), the 1-HER is the most abundant radical in forced oxidation treated wines. Consequently, understanding its evolution kinetic in dry white wines is of great importance.

The interplay between water deficit and nitrogen and potassium nutrition in Vitis vinifera L.

Climate change is expected to provoke an increase in the frequency and intensity of drought events and water scarcity that will have detrimental effects on photosynthesis and plant yield. To sustain an appropriate plant yield under sub-optimal conditions, a common practice is the application of high amounts of fertilizers with negative environmental consequences. The present study aims at evaluating the interplay between water and nutrient availability, namely nitrogen (N) and potassium (K), in two grapevine cultivars with a different sensitivity to water shortage stress. Two-year-old Vitis Vinifera cv. Cabernet Sauvignon and Grenache grapevine plants grafted on SO4 rootstock have been transferred in pots under semi-environmental conditions.

CIEDE2000 colour difference value as a parameter for tracing the ageing process on wood aged spirits

It is quite common nowadays to carry out analyses which allow to control the ageing of spirits that are aged in wood casks. Many control parameters have been previously studied, such as the concentration of different phenolic compounds or the Total Polyphenol Index, in order to better understand the ageing process of wood aged spirits. On the other hand, it is frequent to analyse as a physical parameter the colour of those spirit samples, by stating them as an array of three coordinates from various colour spaces as CIE L*a*b* or CIE L*C*H*.

Predicting oxygen consumption rate by tannins through sweep linear voltammetry and machine learning models

Nowadays, it is well known that oxygen significantly impacts wine quality. The amount of oxygen wine consumes during the winemaking process depends on several factors, such as storage conditions, the number of rackings, the materials used for aging, and the type of closure chosen for bottling.

Which potential for Near Infrared Spectroscopy to characterize rootstock effects on grapevines?

Developing rootstocks adapted to environmental constraints constitutes a key lever for grapevine adaptation to climate change. In this context, Near Infrared Spectroscopy (NIRS) could be used as a high-throughput phenotyping technique to simplify the study of rootstocks in grafted situations. This study is an exploratory analysis to evaluate the potential of NIRS acquired on grafted tissues to reveal rootstock effects as well as the plasticity of combinations of scion/rootstock to better characterize these interactions.
Through the study of 25 combinations (5 scions times 5 rootstocks) in a dedicated experimental vineyard, we showed that NIRS obtained from grafted tissues capture rootstock and scion/rootstock interaction signals, up to 20% of the total variance at specific wavelengths.