Abstract
The grapevine variety plays an important role in wine chain production, thus identifying it is crucial for control activities. However, the specialists responsible for identifying the different varieties, mainly through visual analysis, are disappearing. In this scenario, Deep Learning (DL) classification techniques become a possible solution to handle professionals' scarcity. Nevertheless, previous experiments show that trained classification models use the background information to make decisions, which should be avoided. In this paper, we present a study allowing the assessment of removing background regions from the grapevine images in the improvement classification using DL models. The Xception model is trained with a normal dataset and its segmented version. The Local Interpretable Model-Agnostic Explanations (LIME), Grad-CAM, and Grad-CAM++ approaches are used to visualize the segmentation impact in classification decisions. F1-score of 0.92 and 0.94 were achieved, respectively, for segmented-dataset and normal-dataset trained models. Despite the model trained with the segmented-dataset to achieve a worse performance, the Explainable Artificial Intelligence (XAI) approaches showed that it looks into more reliable regions when making decisions.

Abstract
The grape variety plays an important role in the wine production chain, thus identifying it is crucial for production control. Ampelographers, professionals who identify grape varieties through plant visual analysis, are scarce, and molecular markers are expansive to identify grape varieties on a large scale. In this context, Deep Learning models become an effective way to handle ampelographers scarcity. In this work, we explore the benefit of using deep learning vision transformers architecture relative to conventional CNN to identify 12 grapevine varieties using leaf-centred RGB images acquired in the field. We train an Xception model as a baseline and four different configurations of the ViT_B model. The best model achieved 0.96 of F1-score, outperforming the state-of-the-art convolutional-based model in the used dataset.

Abstract
Crop sustainability is essential for balancing economic development and environmental care, mainly in strong and very competitive regions in the agri-food sector, such as the Region of Murcia in Spain, considered to be the orchard of Europe, despite being a semi-arid area with an important scarcity of fresh water. In this region, farmers apply efficient techniques to minimize supplies and maximize quality and productivity; however, the effects of climate change and the degradation of significant natural environments, such as, the "Mar Menor", the most extent saltwater lagoon of Europe, threatened by resources overexploitation, lead to the search of even better irrigation management techniques to avoid certain effects which could damage the quaternary aquifer connected to such lagoon. This paper describes the Irriman Platform, a system based on Cloud Computing techniques, which includes low-cost wireless data loggers, capable of acquiring data from a wide range of agronomic sensors, and a novel software architecture for safely storing and processing such information, making crop monitoring and irrigation management easier. The proposed platform helps agronomists to optimize irrigation procedures through a usable web-based tool which allows them to elaborate irrigation plans and to evaluate their effectiveness over crops. The system has been deployed in a large number of representative crops, located along near 50,000 ha of the surface, during several phenological cycles. Results demonstrate that the system enables crop monitoring and irrigation optimization, and makes interaction between farmers and agronomists easier.

Abstract
On rugged slopes overlooking the Douro River we find the Alto Douro Wine Region in Portugal, populated by plantations in schist lands of difficult access and mostly manual work. The combined features of this region are a source of motivation to explore remote sensing techniques associated with artificial intelligence. In this paper, a preliminary approach for terrace vineyards detection is presented. This is a key-enabling task towards the achievement of important goals such as multi-temporal crop evaluation and cultures characterization. The proposed methodology consists in the application of a deep learning model (U-net) to detect the terrace vineyards using satellite images dataset acquired with Google Earth Pro. The proposed methodology showed very promising detection capabilities.

Abstract

Abstract