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…

Vitis v. corvina grapes composition and wine sensory profile as affected by different post harvest withering conditions

Context and purpose of the study – In Valpolicella area (Verona – Italy) Vitis vinifera cv. Corvina is the main wine variety to obtain, after grape withering, Amarone wine: this study was carried out in order to compare two different grape dehydration conditions with the aim of verifying the final composition of Corvina dried grapes and the organoleptic profile of corresponding Amarone wine.

Characterization of various groups of pyranoanthocyanins in Merlot red wine

In red wines, anthocyanins evolve during the wine-making process and ageing. They react with other compounds (such as vinylphenols, acetaldehyde, pyruvic acid…) to form a stable family of compounds called pyranoanthocyanins. Furthermore, the oxidation process can modify the anthocyanic profile of a red wine. It is also interesting to evaluate the occurrence of the different subclasses of pyranoanthocyanins and to characterize their chemical properties. The first objective of this study is to evaluate the occurrence of the different groups of pyranoanthocyanins in an oxidised Merlot wine by a centrifugal partition chromatography strategy. The second goal is to evaluate their relative impact in red wines from Bordeaux region by measuring their concentrations.

A novel approach for the identification of new biomarkers of wine consumption in human urine using untargeted metabolomics

Wine is one of the most representative components of Mediterranean diet. Moderate wine intake together with food, has been positively correlated with reduced risk of many chronic diseases. This beneficial effect seems to be ascribed to elevated polyphenolic content of wine [1]. Traditional approaches for the identification of wine biomarkers consumption include targeted metabolomics that focuses on the quantification of well-defined metabolites, losing a valuable information about a massive number of compounds. On the other hand, untargeted metabolomics can disclose a large quantity of signals corresponding to potential biomarkers in a single analysis with high sensitivity and resolution.

Towards a spatial analysis of antique viticultural areas: the case study of Amos (Turkey) and some other places

Interpretation of ancient texts, such as the Amos epigraphic farming leases, questions both locations and spatial extents of the viticultural area, as well as soils, landscapes, cropping methods

Evaluation of wood starch content on bench grafting success rate in grapevine

Since the emergence of phylloxera, grafting has been the most used propagation method in viticulture. Despite all the improvement measures implemented in the nurseries, it is frequent that graft success rates vary depending on the nursery process and scion/rootstock combinations. The reasons behind this unsatisfactory behaviour are still unknown and can be diverse, although carbohydrate reserves might be hypothesised to be crucial, since callus, root, and new tissue formation will be built based on them. In order to identify the effect of carbohydrates on grafting success, nine combinations were established based on the starch content in grapevine scionwoods (cv. Tempranillo clone VN69) and rootstocks cuttings (110 Richter clone 237) used for grafting: Low (L), Medium (M), High (H).